The media and blogs are rife with articles about and condemning the CRTC decision to block GlobeAlive/WIND from winning a wireless telecommunications license. As Michael Geist correctly points out, the blame does not lie with the CRTC. This will require political action to amend the law or, as an interim measure, find some way around it in the near term. I have more faith in the present government acting than I do in the Liberals, when I consider their respective past behaviours with respect to protecting "Canadian-ness" versus open markets.
The incumbent operators are of course pleased with this development -- as well they should be -- as witnessed in this Telus press release and the quote from Rogers in this article. However, I wonder if they will pursue this aggressively when this decision is appealed. It's one thing to gloat but another to directly lobby cabinet. They do not want to be seen to be publicly against competition.
Which brings us back to telecommunications law in this country and the role of the CRTC. Those with a long memory will recall that the Canadian Radio-television and Telecommunications Commission was named the Canadian Radio and Television Commission; same abbreviation (CRTC), but with a expanded mandate. It is no surprise that telecommunications law and regulation mirrors the zealous protection of Canadian content that existed for the media. Ownership matters since that tends to reflect content, business locations, domestic employment and ability of the government to control these companies.
Back when the media and telecommunications sectors were more strongly dominated by a small number, or only one, company -- whether nationally or by region -- there was some legitimate argument for the domestic ownership requirements. Perhaps even more than that, telephone service -- which at one time was almost the entire telecommunications industry -- is rightly held to be an essential public utility. Pre-competition, there was in effect a social contract between the industry and the various governments whereby the industry had to meet a long list of service objectives in return for a virtually guaranteed profit stream.
With competition in place, there is no longer a requirement to protect the incumbents. If one provider fails or unilaterally decides to withdraw from the market, either partially or entirely, the presence of other players helps to ensure continuity of service. Today we have one foot planted in the past and one in the present. We need to make the final step and strongly promote competition so that there is a diminished need for domestic ownership and control over telecommunications providers. As it stands right now, the CRTC's legally correct decision makes no sense when the government's policy is to promoted facilities-based competition. Wireless is the best way to achieve this, and that is what GlobeAlive/WIND will do.
The government will need to act. I predict that they will, and that their decision will be favourable to the GlobeAlive.
Friday, October 30, 2009
Wednesday, October 28, 2009
Trust and the NDA
Dealing with legal agreements is never fun, yet in business it is a daily activity. Most employees rarely have to deal with contracts and other legal documents, leaving that for others and the lawyers they hire. Yet in the high-tech world there is one legal document that is very common: the Non-Disclosure Agreement (NDA). It often goes by other names but let's stick with NDA.
The reason the NDA is so common is that a total product solution for a technology-based company often requires complementary products and services from other companies. This is as true for start-ups as it is for the top names in the sector. These partnerships are initiated by one of the companies or both are directed to do so by a mutual customer. This activity is often delegated to a business development specialist or, for smaller firms, sales. In very early stage companies this may devolve to a technology executive for the simple reason that they may not have a business development or sales group yet. The first step of substantive discussions usually requires an NDA. Later, if all goes well, a contract will be signed to establish the terms of the business relationship.
NDAs are binding legal contracts, as should be obvious if you read through one in detail. They vary quite a bit in length -- I've seen some that fit comfortably within one page, while others run to a dozen pages or more -- but all are intended to achieve one important task: establish procedures, responsibilities and, yes, penalties, to control the necessary sharing of sensitive or proprietary information or technology. The shared intellectual property can range from disclosure of APIs, sharing of source code or pre-release object code, algorithms, customer lists, pricing, and so on.
For those previously unfamiliar with NDAs, it is sometimes assumed that they are a defensive weapon in much the same manner as a suit of armour was for a medieval warrior. With signed NDA in hand, many companies feel free to open up and share their most cherished secrets and the fruits of their labour, believing that the penalties for breaking the terms of the NDA will keep the other party honest. Regrettably, this is often not the case. Here is a recent example that provided the motivation for this blog post.
If you've ever discussed NDAs with lawyers -- which I've done a lot more than I'd like over my career -- you will find that they can be very relaxed about it all (unless they're particularly desperate for a few more billable hours). In contrast, if you are new to the game and are desperate for good advice on whether to sign an NDA or how to draft one, you may be quite anxious since the success or failure of your company may depend on doing everything right. Don't be annoyed with the lawyer's attitude: it's justified. Go ahead and ask why they don't share your concern and you may hear something like this (paraphrased from a discussion I once had with a lawyer along these same lines).
The lawyer will ask you whether you trust the person or company with whom you are negotiating the NDA. This may seem like an odd question since are probably thinking that the NDA is an alternative to trust, countering the risk of dealing with a shady character or a company with a hidden agenda by creating legal consequences for misuse of your valuable intellectual property. It isn't, as is beautifully exemplified by the above-referenced case: if they take the goods and run, regardless of your presumed legal protections, you've already lost.
Sure you can sue, but by then it's too late, it's expensive, the courts can be treacherous and you can be in the right and still lose, and even if you win there may no one you can collect from. It isn't only the small, asset-poor companies you need to watch out for. I have witnessed the very cream of the tech world sign NDAs with the sole objective of getting the goods from start-ups who go on to disclose everything in the hope of winning a top-tier partnership or (though this is only whispered in dark corners) being acquired. For the large corporation it can be far cheaper and faster to exploit what they've learned from you, while knowing that you do not have the resources to successfully pursue a case against a multi-billion dollar corporation in the courts.
The lesson here is that an NDA is absolutely not a substitute for trust. It is a supplement to a trust relationship so that no one makes any mistakes by making clear just what can and cannot be done with disclosed intellectual property. Even so, no matter how strong the level of trust, you should never disclose more than the minimum necessary. Also, disclose gradually in careful steps so that should you realize at some point that you've made a mistake, you can minimize the potential damage. A lawsuit should always be your last option, never the first, and one you almost certainly can't afford.
As when selecting a romantic partner or a building contractor, ask around discretely before walking down the aisle or tearing down the walls. Also, when you do sign an NDA, treat your new partner's intellectual property with the same care and diligence with which you wish they would treat yours. That way when the next deal comes around and the other party checks out your reputation, your existing partners will give the thumbs up sign.
The reason the NDA is so common is that a total product solution for a technology-based company often requires complementary products and services from other companies. This is as true for start-ups as it is for the top names in the sector. These partnerships are initiated by one of the companies or both are directed to do so by a mutual customer. This activity is often delegated to a business development specialist or, for smaller firms, sales. In very early stage companies this may devolve to a technology executive for the simple reason that they may not have a business development or sales group yet. The first step of substantive discussions usually requires an NDA. Later, if all goes well, a contract will be signed to establish the terms of the business relationship.
NDAs are binding legal contracts, as should be obvious if you read through one in detail. They vary quite a bit in length -- I've seen some that fit comfortably within one page, while others run to a dozen pages or more -- but all are intended to achieve one important task: establish procedures, responsibilities and, yes, penalties, to control the necessary sharing of sensitive or proprietary information or technology. The shared intellectual property can range from disclosure of APIs, sharing of source code or pre-release object code, algorithms, customer lists, pricing, and so on.
For those previously unfamiliar with NDAs, it is sometimes assumed that they are a defensive weapon in much the same manner as a suit of armour was for a medieval warrior. With signed NDA in hand, many companies feel free to open up and share their most cherished secrets and the fruits of their labour, believing that the penalties for breaking the terms of the NDA will keep the other party honest. Regrettably, this is often not the case. Here is a recent example that provided the motivation for this blog post.
If you've ever discussed NDAs with lawyers -- which I've done a lot more than I'd like over my career -- you will find that they can be very relaxed about it all (unless they're particularly desperate for a few more billable hours). In contrast, if you are new to the game and are desperate for good advice on whether to sign an NDA or how to draft one, you may be quite anxious since the success or failure of your company may depend on doing everything right. Don't be annoyed with the lawyer's attitude: it's justified. Go ahead and ask why they don't share your concern and you may hear something like this (paraphrased from a discussion I once had with a lawyer along these same lines).
The lawyer will ask you whether you trust the person or company with whom you are negotiating the NDA. This may seem like an odd question since are probably thinking that the NDA is an alternative to trust, countering the risk of dealing with a shady character or a company with a hidden agenda by creating legal consequences for misuse of your valuable intellectual property. It isn't, as is beautifully exemplified by the above-referenced case: if they take the goods and run, regardless of your presumed legal protections, you've already lost.
Sure you can sue, but by then it's too late, it's expensive, the courts can be treacherous and you can be in the right and still lose, and even if you win there may no one you can collect from. It isn't only the small, asset-poor companies you need to watch out for. I have witnessed the very cream of the tech world sign NDAs with the sole objective of getting the goods from start-ups who go on to disclose everything in the hope of winning a top-tier partnership or (though this is only whispered in dark corners) being acquired. For the large corporation it can be far cheaper and faster to exploit what they've learned from you, while knowing that you do not have the resources to successfully pursue a case against a multi-billion dollar corporation in the courts.
The lesson here is that an NDA is absolutely not a substitute for trust. It is a supplement to a trust relationship so that no one makes any mistakes by making clear just what can and cannot be done with disclosed intellectual property. Even so, no matter how strong the level of trust, you should never disclose more than the minimum necessary. Also, disclose gradually in careful steps so that should you realize at some point that you've made a mistake, you can minimize the potential damage. A lawsuit should always be your last option, never the first, and one you almost certainly can't afford.
As when selecting a romantic partner or a building contractor, ask around discretely before walking down the aisle or tearing down the walls. Also, when you do sign an NDA, treat your new partner's intellectual property with the same care and diligence with which you wish they would treat yours. That way when the next deal comes around and the other party checks out your reputation, your existing partners will give the thumbs up sign.
Monday, October 26, 2009
Head of State and Republicanism
Like so many peculiar Canadian myths, the one regarding the monarchy continues to baffle me. Whenever a media commentator discusses the question, whether it be pro or con, it warms the blood of many otherwise complacent folk. This article is the latest one that caught my attention.
The reason the question of the monarchy baffles me is that it is a settled question, regardless of the seeming controversy. By saying that I may have in turn baffled you! Let me explain what I mean.
Canada is not a monarchy. It is a republic. Yes, it is true that there is a piece of parchment somewhere that says we are -- as schoolchildren are taught -- a constitutional monarchy. There are also the visible signs of the monarchy such as the Governor General, the monarch's image and Latin caption on our money, occasional royal visits, and so on. This is the obvious stuff that blinds some into believing that we do have a monarch. We don't. Starting with so-called responsible government well before Confederation and up to the ultimate act of making the Supreme Court the highest court of appeal about three decades ago, we have moved slowly but steadily from a monarchy to a republic. But as of three decades ago, Canada is a fully-operational republic. No, not on paper (sorry, parchment), but in the reality of how power is wielded.
The monarch has no regal power in this country. None. Try to image if you can Queen Elizabeth actually attempting to influence, let allow propose or quash, acts of Parliament, or, going even further, dissolve Parliament. If she or her successor were silly enough to try any of these things you can bet that we would very quickly do what's needed to become a republic on paper (or parchment) as well. The government wouldn't stand for it and neither would the "loyal" opposition or us, the citizens. If we ask the same question of the Governor General, it is less clear, as we discovered earlier this year. The difference is that, while usually a purely ceremonial role, the role of GG does have power, but that power originates with the government -- a government elected by the people -- that nominates the sole candidate for the position. This is just like one style of republic where the President is appointed by the government or elected by Parliament; another other way is to elect the President by popular vote, though this step is often skipped where the role is not one of chief executive.
The role of GG or President is not entirely ceremonial since there are cases where parliamentary deadlocks must be broken (e.g. dismiss the government or ask another party to try to form one), but also can serve as a check to a broadly-unpopular act of Parliament when there is no other body to vet legislation. In Canada's case, the Senate is almost entirely toothless in this regard. I suggest that this one item is where the power of a GG contrasts with that of a ceremonial President: the GG does not dare to block legislation, whereas a President might. The difference is small though necessary but, importantly, has nothing whatever to do with the monarch.
Whether we take the step to formally abolish the monarchy is of no importance to me. It's enough that I know that Canada is, where it matters most -- formal power -- a republic. The imprint on our coins doesn't affect this reality.
The reason the question of the monarchy baffles me is that it is a settled question, regardless of the seeming controversy. By saying that I may have in turn baffled you! Let me explain what I mean.
Canada is not a monarchy. It is a republic. Yes, it is true that there is a piece of parchment somewhere that says we are -- as schoolchildren are taught -- a constitutional monarchy. There are also the visible signs of the monarchy such as the Governor General, the monarch's image and Latin caption on our money, occasional royal visits, and so on. This is the obvious stuff that blinds some into believing that we do have a monarch. We don't. Starting with so-called responsible government well before Confederation and up to the ultimate act of making the Supreme Court the highest court of appeal about three decades ago, we have moved slowly but steadily from a monarchy to a republic. But as of three decades ago, Canada is a fully-operational republic. No, not on paper (sorry, parchment), but in the reality of how power is wielded.
The monarch has no regal power in this country. None. Try to image if you can Queen Elizabeth actually attempting to influence, let allow propose or quash, acts of Parliament, or, going even further, dissolve Parliament. If she or her successor were silly enough to try any of these things you can bet that we would very quickly do what's needed to become a republic on paper (or parchment) as well. The government wouldn't stand for it and neither would the "loyal" opposition or us, the citizens. If we ask the same question of the Governor General, it is less clear, as we discovered earlier this year. The difference is that, while usually a purely ceremonial role, the role of GG does have power, but that power originates with the government -- a government elected by the people -- that nominates the sole candidate for the position. This is just like one style of republic where the President is appointed by the government or elected by Parliament; another other way is to elect the President by popular vote, though this step is often skipped where the role is not one of chief executive.
The role of GG or President is not entirely ceremonial since there are cases where parliamentary deadlocks must be broken (e.g. dismiss the government or ask another party to try to form one), but also can serve as a check to a broadly-unpopular act of Parliament when there is no other body to vet legislation. In Canada's case, the Senate is almost entirely toothless in this regard. I suggest that this one item is where the power of a GG contrasts with that of a ceremonial President: the GG does not dare to block legislation, whereas a President might. The difference is small though necessary but, importantly, has nothing whatever to do with the monarch.
Whether we take the step to formally abolish the monarchy is of no importance to me. It's enough that I know that Canada is, where it matters most -- formal power -- a republic. The imprint on our coins doesn't affect this reality.
Thursday, October 22, 2009
CRTC Decision 2009-657 on Traffic Management
It's only the second day since the CRTC issued its ruling on ISP traffic management and there is already a large number of articles (here and here, etc.) -- summaries of the ruling and analyses -- circulating in print and on the internet. There is little point in having me do the same, so I won't repeat what others have already done. Instead I want to look at just how this ruling benefits the large ISPs -- the major carriers against whom the complaints were originally made -- beyond what has already been mentioned. I believe there is a reason why these companies are pretty happy with the outcome:
Apart from the notification requirement, the carriers are in total control of their actions. What I mean is, the CRTC is leaving every judgment call and technical determination of impact on retail and wholesale customer to the carriers; the CRTC has not injected themselves anywhere into the process whereby the carriers must seek approval before acting as they see fit. If you read the decision yourself, you may very well believe that I am wrong about this since the requirements the CRTC lists are, by and large, neutral to positive for the carriers' customers. The only mention I've come across that shows any level of doubt is that it is up to the carriers' customers to call attention to any inappropriate traffic management behaviour, which is very difficult to prove from outside the network. I believe the impact is greater than this would indicate.
First, notification is only a requirement if the carrier is instituting traffic management that is more restrictive than what is extant (paragraph 80). The only judge of this is the carrier itself, and so they can choose to not disclose what they've done (paragraph 84). Further, the criteria to measure restriction are exceedingly loose, apart from VoIP and media streaming (paragraphs 126 and 127) which requires pre-approval to throttle or block. This leaves the carrier free to argue, whenever the change is noticed and they are questioned about it, that what they are doing is not more restrictive than what was done before, just different. When conflict inevitably arises, as with the present issue, they will almost certainly continue on with their new traffic management practice unless and until their customers convince the CRTC to act against them. As we've seen, this can take a very long time, during which the carrier can do as they please.
One surprise is that they ruled wireless (mobile) data operators must conform the same traffic management behaviours ruling (paragraph 116). Although it will suffer from the same loopholes as cable and DSL internet access, it is something positive for consumers. This is an interesting development in light of the FCC's proposed rules out today (October 22) on the same matter: wireless network neutrality.
CRTC intervention may be even less likely in future, or at least longer in coming, with their direction to the industry to cooperate and negotiate over points of dispute (paragraph 76). This is not a Commission that intends to be interventionist, despite offers to hear complaints in certain cases (paragraphs 127 and 128).
While it is certainly true that there are political arguments to be made as to why the CRTC is acting as they are, there is an economic reason as well. Reading paragraph 35, the CRTC is consistent in not wanting (if you like) to kill the goose that is laying the golden eggs: it's the carriers that are making the investment and taking the business risk of building the internet infrastructure in Canada. ISPs that make use of GAS (DSL wholesale) do not, although they so provide a source of price and service competitive balance in an otherwise near-monopoly market.
My opinion is that they will continue to regulate with a light hand, and in so doing they will tolerate modestly discriminatory behaviour, believing that if the carriers' balance sheets remain strong, investment in their networks will continue. In contrast, they are not (yet) willing to bet on new entrants or existing small players to take a major role in building out infrastructure. Whether you consider this good or bad often depends on where your own interest lie. There is no absolute right and wrong in this matter despite attempts by some to colour it that way.
BCE said in a statement that it thinks the decision is a good one and that its "existing Internet traffic management practices are already compliant with it."I suspect these companies are happier with the ruling than is apparent in their public statements; they do not want to be seen as gloating. Let me now go through my reasoning why this is a potentially big win for these companies.
Michael Hennessy, senior vice-president of regulatory and government affairs at Telus Corp (T.TO), said the communications company doe not currently throttle traffic. However, it does employ some general caps on bandwidth usage.
"There are growing concerns about congestion," he said, adding the CRTC's decision is "very good and very fair", and that continued network investments and consumption-based pricing are among ways to address heavy traffic volumes.
Apart from the notification requirement, the carriers are in total control of their actions. What I mean is, the CRTC is leaving every judgment call and technical determination of impact on retail and wholesale customer to the carriers; the CRTC has not injected themselves anywhere into the process whereby the carriers must seek approval before acting as they see fit. If you read the decision yourself, you may very well believe that I am wrong about this since the requirements the CRTC lists are, by and large, neutral to positive for the carriers' customers. The only mention I've come across that shows any level of doubt is that it is up to the carriers' customers to call attention to any inappropriate traffic management behaviour, which is very difficult to prove from outside the network. I believe the impact is greater than this would indicate.
First, notification is only a requirement if the carrier is instituting traffic management that is more restrictive than what is extant (paragraph 80). The only judge of this is the carrier itself, and so they can choose to not disclose what they've done (paragraph 84). Further, the criteria to measure restriction are exceedingly loose, apart from VoIP and media streaming (paragraphs 126 and 127) which requires pre-approval to throttle or block. This leaves the carrier free to argue, whenever the change is noticed and they are questioned about it, that what they are doing is not more restrictive than what was done before, just different. When conflict inevitably arises, as with the present issue, they will almost certainly continue on with their new traffic management practice unless and until their customers convince the CRTC to act against them. As we've seen, this can take a very long time, during which the carrier can do as they please.
One surprise is that they ruled wireless (mobile) data operators must conform the same traffic management behaviours ruling (paragraph 116). Although it will suffer from the same loopholes as cable and DSL internet access, it is something positive for consumers. This is an interesting development in light of the FCC's proposed rules out today (October 22) on the same matter: wireless network neutrality.
CRTC intervention may be even less likely in future, or at least longer in coming, with their direction to the industry to cooperate and negotiate over points of dispute (paragraph 76). This is not a Commission that intends to be interventionist, despite offers to hear complaints in certain cases (paragraphs 127 and 128).
While it is certainly true that there are political arguments to be made as to why the CRTC is acting as they are, there is an economic reason as well. Reading paragraph 35, the CRTC is consistent in not wanting (if you like) to kill the goose that is laying the golden eggs: it's the carriers that are making the investment and taking the business risk of building the internet infrastructure in Canada. ISPs that make use of GAS (DSL wholesale) do not, although they so provide a source of price and service competitive balance in an otherwise near-monopoly market.
My opinion is that they will continue to regulate with a light hand, and in so doing they will tolerate modestly discriminatory behaviour, believing that if the carriers' balance sheets remain strong, investment in their networks will continue. In contrast, they are not (yet) willing to bet on new entrants or existing small players to take a major role in building out infrastructure. Whether you consider this good or bad often depends on where your own interest lie. There is no absolute right and wrong in this matter despite attempts by some to colour it that way.
Wednesday, October 21, 2009
Google Ultimatum
In a post a few weeks ago I concluded that the FCC is very likely to rule that Google Voice is subject to regulation under US telecommunications law. Since then, the FCC has taken the next step of gathering information, including requesting information from Google by means of this letter, a letter than while widely referenced, appears to have received little direct analysis by the trade media and bloggers.
Alongside this, there is the sideshow of AT&T and others attempting to sway the FCC -- by attempting to stir up public outrage against Google -- but which are merely distractions from the main event: the legal determination of whether Google Voice should be subject to federal telecommunications regulation. While opinions and lobbying can have an impact, none of this should cloud our view of what is transpiring.
Unlike in Canada, the federal telecommunications regulator is tightly constrained by federal statutes in what they can and must do in the performance of their duties. If you were to compare US and Canadian law, one important difference is that US law goes much deeper into dictating carrier operations and, therefore, how the FCC must regulate those carriers. That is one reason why many FCC staffers are lawyers, as are the industry reps who interact with the FCC, and not incidentally, why every FCC action (and industry reply) is full of legal footnotes, just like in this letter to Google.
In Canada, the law specifies more general guidelines, leaving the CRTC free to develop telecommunications policy. A further consequence is that political direction from the government of the day is more likely in Canada; in the US, Congress and the Administration are somewhat stymied by the more detailed laws -- which they created! -- that rule the FCC's actions.
In light of this, let's look at the FCC's letter more closely so that we can understand how this public process is unfolding, and what the FCC may really be up to. I'll do this by quoting and commenting on brief extracts from the letter.
All in all, Google is in a quandary. Their answers will almost certainly result in a determination that Google Voice is a telecommunications service, and therefore they will no longer be permitted to block calls. This will cause some havoc with Google's business operations. In particular, they will at a minimum be required to operate Google Voice under accounting separation from the rest of Google's business and place this segregated operation under regulatory oversight. If they are very unlucky -- although I don't believe the FCC will go this far -- Google could be required to structurally separate Google Voice, operating it as a subsidiary.
With their reply to the FCC due October 28, it won't be very long -- probably by year end -- until we know how the FCC will rule. If it goes against Google, expect an appeal: first to the FCC, and then to the courts. This matter could take some time to come to an ultimate end. The result is important to not only Google, but to every operator of similar services and VoIP providers.
Alongside this, there is the sideshow of AT&T and others attempting to sway the FCC -- by attempting to stir up public outrage against Google -- but which are merely distractions from the main event: the legal determination of whether Google Voice should be subject to federal telecommunications regulation. While opinions and lobbying can have an impact, none of this should cloud our view of what is transpiring.
Unlike in Canada, the federal telecommunications regulator is tightly constrained by federal statutes in what they can and must do in the performance of their duties. If you were to compare US and Canadian law, one important difference is that US law goes much deeper into dictating carrier operations and, therefore, how the FCC must regulate those carriers. That is one reason why many FCC staffers are lawyers, as are the industry reps who interact with the FCC, and not incidentally, why every FCC action (and industry reply) is full of legal footnotes, just like in this letter to Google.
In Canada, the law specifies more general guidelines, leaving the CRTC free to develop telecommunications policy. A further consequence is that political direction from the government of the day is more likely in Canada; in the US, Congress and the Administration are somewhat stymied by the more detailed laws -- which they created! -- that rule the FCC's actions.
In light of this, let's look at the FCC's letter more closely so that we can understand how this public process is unfolding, and what the FCC may really be up to. I'll do this by quoting and commenting on brief extracts from the letter.
RE: Google Voice Calling RestrictionsRight up front we have the FCC's focus on a critical obligation of common carriers: non-discrimination. This is a long-standing requirement that both protects the common carrier from prosecution when their services are used in the performance of a crime and ensures universality of this important utility: where anyone can call anyone else. Importantly, the rule is also used to prevent carriers from gaming the FCC and state PUC-mandated termination fees that benefit rural local exchange carriers (LEC), which are intended to subsidize service prices in high-cost (low density) locales. This is a politically-sensitive topic for all levels of government.
...please provide answers to the following questions by close of business on Wednesday, October 28, 2009.At present, since Google has not been determined to be a carrier, the FCC cannot demand this data, but must request it. Google is not legally obliged to respond, but refusal would stain their image and would in any case not stop the FCC from proceeding with information gleaned from other sources, including their antagonists, such as AT&T.
...please describe how the Google Voice call is routed...Does Google contract with third parties to obtain inputs for its Google Voice service, such as access to telephone numbers, transmission of telephone calls, and interconnection with local telephone networks?In my previous post, I described how Google Voice is structured as a shell around a (partner) carrier's telephone network. FCC, of course, already knows this, but needs Google to state it explicitly. Which brings us to the following item:
Does the Google Voice service compete with any services classified as “telecommunications services” under the Act? Is Google Voice a reseller of “telecommunications services?”The shell structure belies any Google claim that Google Voice is not a telecommunications service. Here the FCC is anticipating and likely to attack Google claims that the service is an related software or web application, or something like a PBX or other business phone which connects to a telecommunications service, and is therefore not a telecommunications service itself. The FCC's questions lead me to believe that they are ready to reject these avenues of defense.
For each functionality for which calls to particular telephone numbers are restricted, please describe the technological means by which those restrictions are implemented.I suspect that what the FCC wants to know is if the blocking is performed in Google's own software or in their partner's network. In my previous post I speculated that the former is true; if so, they are performing a form of telephony routing, and this could hurt their case. If it is the latter, expect the FCC to dig deeper to understand how their partner decides to do the blocking: on their own or in cahoots with Google. The latter is almost certainly true since the FCC has successfully been clamping down on carriers who have recently been blocking calls to selected numbers. What I expect to hear from Google is that the carrier is performing this function on Google's request -- Google is their customer -- and not at the carrier's own initiative.
To what extent, if any, does Google charge for any of these services? Does Google intend to charge at some point for the service? How does Google currently pay for the service?If there's money involved, that's a business, and the FCC is likely to use that as a basis to rule that Google is in the telecommunications business. While it's true that Google does not currently charge users directly for Google Voice, there is a flow of dollars. This may include ads on Google Voice's web views (even if not done today), but certainly does include payment to their carrier partner for call origination (points-of-presence and Caller Id setting) and incoming call routing (redirection from the Google Voice number to a physical phone number). The fact that Google is operating the service at a loss is not likely to be a mitigating factor since many carriers don't earn a profit.
...explain specifically what is meant by “invitation-only.” How many users of Google Voice are there at this time? Are there any plans to offer Google Voice on other than an invitation-only basis?Here it appears that the FCC intends to puncture any defense based on Google Voice not being a generally-available public offer. This goes back to the previous discussion regarding non-discrimination. Is the invitation a marketing ploy or perhaps a means to manage service expansion? Either way, the end point is general availability, and similar Google examples such as Gmail can be referenced by the FCC to buttress this angle of attack.
If Google requests that any information or documents responsive to this letter be treated in a confidential manner...This is a carrot to ensure Google is forthcoming with the information being requested. They are offering a blanket protection at this stage of the game, however the threshold for confidentiality will get higher should the FCC deem that Google Voice is a telecommunications service. As a common carrier, regular reporting of data, including financial data, will become mandatory, and some portion of it will become public.
All in all, Google is in a quandary. Their answers will almost certainly result in a determination that Google Voice is a telecommunications service, and therefore they will no longer be permitted to block calls. This will cause some havoc with Google's business operations. In particular, they will at a minimum be required to operate Google Voice under accounting separation from the rest of Google's business and place this segregated operation under regulatory oversight. If they are very unlucky -- although I don't believe the FCC will go this far -- Google could be required to structurally separate Google Voice, operating it as a subsidiary.
With their reply to the FCC due October 28, it won't be very long -- probably by year end -- until we know how the FCC will rule. If it goes against Google, expect an appeal: first to the FCC, and then to the courts. This matter could take some time to come to an ultimate end. The result is important to not only Google, but to every operator of similar services and VoIP providers.
Wednesday, October 14, 2009
The Myth of Dollar Parity
All the major media are bombarding us daily with stories about the approaching so-called parity between the Canadian and US dollar exchange rate. Now it's a horse race, with people even giving probabilities as to if and when this magical milestone will be achieved:
Why does this matter? I mean, aside from the perverse pride many Canadians feel -- due, it seems, to a silly need to give Americans some comeuppance -- when this meaningless arithmetic coincidence occurs, does it have any operational impact on us and our economy. Sure, it does matter to our economic health that there is exchange rate volatility and a shifting trade imbalance with the United States, but this one particular point on the exchange rate continuum. Most emphatically, no!
Let's say, for the sake of argument, that rather than being denominated in units we call dollars, the Canadian currency unit will change overnight to the drachma, at the rate of 2 drachmas to the old dollar. The numerical value of all prices, stock quotes, etc. will now be precisely twice what it was the day before. Further the exchange rate with the US dollar becomes 0.50d (where d is drachmas). The question, what has changed?
The answer is, apart from a loss of so-called parity, absolutely nothing. It should be clear from this that parity has no operational impact. About the only impact of parity is psychological: Canadians start feeling better about themselves. Now that's a ridiculous reason for anyone to focus on currency parity.
About the only practical benefit I get from parity is the easy comparison of prices, those north and south of the border, of the same product or commodity. That's it. I suspect the media is spending so much ink (or pixels) on this topic is that there are diminishing hopes for a federal election: the currency is apparently their second choice of a horse race to latch on to.
"Canada's dollar has a 68 per cent probability of reaching parity with the greenback by year-end, up from 61 per cent yesterday, according to implied volatility from options trading monitored by Bloomberg."With some software help, I counted 10 occurrences of the word parity in the quoted article.
Why does this matter? I mean, aside from the perverse pride many Canadians feel -- due, it seems, to a silly need to give Americans some comeuppance -- when this meaningless arithmetic coincidence occurs, does it have any operational impact on us and our economy. Sure, it does matter to our economic health that there is exchange rate volatility and a shifting trade imbalance with the United States, but this one particular point on the exchange rate continuum. Most emphatically, no!
Let's say, for the sake of argument, that rather than being denominated in units we call dollars, the Canadian currency unit will change overnight to the drachma, at the rate of 2 drachmas to the old dollar. The numerical value of all prices, stock quotes, etc. will now be precisely twice what it was the day before. Further the exchange rate with the US dollar becomes 0.50d (where d is drachmas). The question, what has changed?
The answer is, apart from a loss of so-called parity, absolutely nothing. It should be clear from this that parity has no operational impact. About the only impact of parity is psychological: Canadians start feeling better about themselves. Now that's a ridiculous reason for anyone to focus on currency parity.
About the only practical benefit I get from parity is the easy comparison of prices, those north and south of the border, of the same product or commodity. That's it. I suspect the media is spending so much ink (or pixels) on this topic is that there are diminishing hopes for a federal election: the currency is apparently their second choice of a horse race to latch on to.
VoIP, Mobile and Canada - an Update
In an earlier article I mentioned that, while technically feasible, there are several factors that make VoIP largely unusable over mobile (wireless) carriers' data networks in Canada. Some of these factors are now being overcome in the US, sooner than I expected. Let's briefly review these issues from a Canadian perspective:
- Competitive pressures: Nothing has changed as yet, with new licensees not ready to make an appearance on the Canadian market for some months yet, at the very least.
- Receiving incoming calls: Smart phones can easily run VoIP, however being available for incoming calls has a deleterious impact on the battery life, among other technical hurdles. The first sign that this barrier is coming down is Apple's announcement of iPhone OS 3.0 which enables an application to receive push notifications (APNS). It isn't reliably real-time, but it is a step in the right direction.
- VoIP application availability: With AT&T prepared to allow VoIP apps like Skype and Vonage onto its 3G data network, Apple has less reason to block any VoIP app. While they are still delaying Google Voice, since it is not a VoIP app it is an unrelated issue. Of course, VoIP apps can be used on Android phones.
- Application filtering: This is the mobile version of the net neutrality public policy issue. Carriers have the technical ability to filter VoIP apps (communications protocols in the data channel), and they frequently do this to protect their voice business and (controversially) to protect overall service quality on their data networks. Again, this is where competition is needed to loosen restrictions in Canada since the CRTC cannot be relied upon to act.
- Data plan pricing: This remains expensive in Canada in comparison to other countries, and VoIP requires more affordable pricing if VoIP is to become a competitive alternative to the carrier's own voice service.
Tuesday, October 13, 2009
Categories: Blogs vs. News
There is an ongoing debate these days about whether bloggers are journalists and whether blogs are news sources. This is further muddied by journalists who blog, bloggers that don't merely refer to media news sources but do their own investigative reporting, and news sites that are little more than aggregated, filtered and organized collections of blog postings. This is all very interesting, but not an issue that I deeply care about. My problem is more mundane: how can I best organize my browser bookmarks.
There was a time when I clearly viewed blogs and traditional media sources are distinct things. To deal with that, it was enough to create separate bookmark folders for blogs and news sites, further categorized by my business and personal interests. As time went on, this distinction came to make less and less sense. My bookmarking behaviour in response to this was a bit haphazard, resulting in an unpredictable mix of blog and news sites within various folders and categories. It got so bad that I could sometimes not predict in which folder I might find a particular bookmark or, for new bookmarks, decide where to file them. This is confusing enough that I often see the same thing in Yahoo! stock summaries, where they, too, have separate categories for news and blogs pertaining to a stock, yet it is often unpredictable in which category any particular article will appear.
This past week I decided (as I will explain) that the distinction was no longer sufficient to justify keeping these bookmarks separate. I reorganized my bookmarks by interest category, and within those categories I order all bookmarks by importance or frequency of use. That is, blogs are no longer a category. After having made this change, my bookmarks seem far more natural and easier to navigate. Since I use Firefox, I can circumvent categories completely for bookmarks and RSS feeds that are of high importance, by placing them on the bookmarks toolbar for one-click access.
Apart from making my time in front on my computer(s) a bit less stressful, this matter evokes a deeper question about just what it is that sets blogs apart from the web properties of old and new media enterprises. Blogs, especially those operated by an individual, are typically dedicated to a particular topic, with posts shown in reverse chronological order. While there is the ability to organize posts by blogger-defined categories, the more common technique used is tags: search on a tag, and you see the posts within that category. Tags are also handy for blog search engines and aggregators, so that they only pick up pertinent articles, or at least better organize them. In contrast, news sites use editorial control (human or automated) to categorize articles and present these categorized lists on their front pages.
Of course, it isn't quite as simple as I've presented. News sites, including many blog aggregators, provide both a descending chronological list of the most recent or important articles in addition to categories (example: Globe and Mail). News sites, like blogs, also provide RSS feeds that are similarly presented in reverse chronological order. I believe what this demonstrates is that blogs and news sites are increasing converging to a small set of presentation methods, and that they are doing so because that is what works best for readers. As the blogs vs. news matter continues to evolve, I expect that this convergence will continue.
Most importantly, this means my bookmarking woes should eventually come to an end. This assumes that my interests and, therefore, bookmark categories, remain stable. Unfortunately (or, fortunately) my interests change over time, and so I can never completely escape periodic reorganizations.
There was a time when I clearly viewed blogs and traditional media sources are distinct things. To deal with that, it was enough to create separate bookmark folders for blogs and news sites, further categorized by my business and personal interests. As time went on, this distinction came to make less and less sense. My bookmarking behaviour in response to this was a bit haphazard, resulting in an unpredictable mix of blog and news sites within various folders and categories. It got so bad that I could sometimes not predict in which folder I might find a particular bookmark or, for new bookmarks, decide where to file them. This is confusing enough that I often see the same thing in Yahoo! stock summaries, where they, too, have separate categories for news and blogs pertaining to a stock, yet it is often unpredictable in which category any particular article will appear.
This past week I decided (as I will explain) that the distinction was no longer sufficient to justify keeping these bookmarks separate. I reorganized my bookmarks by interest category, and within those categories I order all bookmarks by importance or frequency of use. That is, blogs are no longer a category. After having made this change, my bookmarks seem far more natural and easier to navigate. Since I use Firefox, I can circumvent categories completely for bookmarks and RSS feeds that are of high importance, by placing them on the bookmarks toolbar for one-click access.
Apart from making my time in front on my computer(s) a bit less stressful, this matter evokes a deeper question about just what it is that sets blogs apart from the web properties of old and new media enterprises. Blogs, especially those operated by an individual, are typically dedicated to a particular topic, with posts shown in reverse chronological order. While there is the ability to organize posts by blogger-defined categories, the more common technique used is tags: search on a tag, and you see the posts within that category. Tags are also handy for blog search engines and aggregators, so that they only pick up pertinent articles, or at least better organize them. In contrast, news sites use editorial control (human or automated) to categorize articles and present these categorized lists on their front pages.
Of course, it isn't quite as simple as I've presented. News sites, including many blog aggregators, provide both a descending chronological list of the most recent or important articles in addition to categories (example: Globe and Mail). News sites, like blogs, also provide RSS feeds that are similarly presented in reverse chronological order. I believe what this demonstrates is that blogs and news sites are increasing converging to a small set of presentation methods, and that they are doing so because that is what works best for readers. As the blogs vs. news matter continues to evolve, I expect that this convergence will continue.
Most importantly, this means my bookmarking woes should eventually come to an end. This assumes that my interests and, therefore, bookmark categories, remain stable. Unfortunately (or, fortunately) my interests change over time, and so I can never completely escape periodic reorganizations.
Thursday, October 8, 2009
Election Deferred
It increasingly appears that a federal election won't occur this fall. Like most people, I'm ok with that. In a post back in August, I suggested that the Liberals would like an election now so that they can be in office when the economy recovers:
In light of the Liberal bungling -- no credible reason presented for defeating the government -- it is no surprise they are lagging in the polls. Harper has now taken the high road in addressing the election question by claiming that the poll numbers mean there is no likelihood of an election, and everyone can relax. This is the perfect strategy for the Conservatives because now they can succeed at the supposed Liberal strategy I had outlined in my previous post: they want to be the party in power when the economy recovers so that they can claim the credit.
"Whatever justification Ignatieff ultimately chooses to promote, I do not believe it will be the real reason. I expect that the true, and hidden, reason will be the same one why he has chosen to support the government until now: the economy. More specifically, they want to stain the Conservatives with the economic downturn and recession, while making sure they are in government when the economy turns around..."This is pretty much how Ignatieff played his hand, although he bungled it badly. By strictly holding to his intention to only reveal his platform after an election is called, he had nothing to engage the electorate to justify triggering an election. Harper took advantage of that by playing the EI card, making the Liberals look ridiculous in their opposition to a reasonable set of EI reforms. They then snared the NDP into supporting the government, a party that is particularly fearful of an election at this time since they know they would get wiped out, electorally and financially.
In light of the Liberal bungling -- no credible reason presented for defeating the government -- it is no surprise they are lagging in the polls. Harper has now taken the high road in addressing the election question by claiming that the poll numbers mean there is no likelihood of an election, and everyone can relax. This is the perfect strategy for the Conservatives because now they can succeed at the supposed Liberal strategy I had outlined in my previous post: they want to be the party in power when the economy recovers so that they can claim the credit.
"The truth is that neither the Conservatives nor the Liberals are responsible for either the recession or the coming economic resurgence..."This remains as true as when I said it in August. With increasing signs of at least a tepid recovery in 2010, the Conservatives are well on the way to winning a majority should an election be called any time during 2010. They will get (unearned) credit for the economic recovery -- assuming we all forget about the massive debt burden, which is entirely possible -- and for steering the country away from an unwanted election, while looking extraordinarily competent in comparison to the Liberals and NDP.
Wednesday, October 7, 2009
App Developers & App Stores: Misaligned Objectives
There was some interesting news last week about a new investment fund for early stage stage ventures focused on mobile applications. Any early stage funds availability is good news for technology entrepreneurs in the Ottawa region, and it certainly caught my attention considering my own interest in mobile apps and Android. Regardless of how the fund will operate, any company hoping to tap into this source of capital will need a credible business plan. For mobile app developers, this is no easy feat.
A large majority of apps on iPhone (85,000+) and Android (10,000+) are either free or are not generating enough revenue to pay the developer's home utility bills, let alone achieving profitability. The odds for success are much better than lottery tickets, but there is a lot more to being successful than coding like mad for a few weeks and then praying for the money to come rolling in. Tapping into the consumer market is particularly fraught with peril since it is so very hard to predict how the market will respond. This is anathema to developing a workable business plan unless there is already some market success for a core part of the planned functionality from which revenue extrapolation can be reliably attempted. Going after the enterprise market is safer since the prospective customers will buy if they have a good business case for a product, and that determines the app's revenue prospects. Setting yourself up so that you're subject to the whims of consumer tastes of the moment will not endear you to investors.
What we need to consider is just what the business model actually entails when selling mobile apps, regardless of the target market. The majority of mobile apps are distributed through app stores run by Apple, Google, Microsoft, RIM and Palm, for apps that are designed for their own brands of mobile OS platforms. There are third-party stores and possibilities for carrier stores (although Verizon, in their announcement this week, surprisingly to some, claimed their Android phones will rely on Google Android Market), but we can discount these parallel distribution channels for the present. Because of developers' reliance on these app stores for marketing, distribution and payment processing, for which they pay a price (typically 30% of sales revenue, split between the app store operator and the carrier in most cases), it is important to assess whether this "partnership" is to the benefit of both parties.
For the platform vendor (which is the same as the device vendor in the cases of Apple, RIM and Palm), there are two key benefits of an app store:
With regard to the second point -- revenue -- the dynamics are far more interesting, and telling. I have seen a number of contradictory stories of how much money Apple is deriving for their app store (either very large or very small), and how much of that is shared with carriers. Even if the true number is toward the high end, the reported figures still pale in comparison to the revenue derived from iPhone and iPod. That is, it's a nice pile of money but not enough to get too excited about for a corporation the size of Apple. In the case of Google, the amount must be far lower since we know that the majority of the 30% fee goes to the carriers, and there is much lower success rate of paid apps on the Android Market. (Google, as I'll discuss, has little incentive to do better.)
Putting these facts and projections together, I conclude that the primary objective is the first point: increasing the device value proposition by providing users with a large catalogue of applications, both free and paid. The problem for app developers is that Apple and Google have already achieved this objective, and therefore have little incentive to make any further attempts to help app developers to succeed. App developers would obviously prefer that their so-called partners' primary objective were to maximize app revenue, but that is not the present situation. We should expect the same to hold true for RIM, Microsoft and Palm when or if they pass the same app catalogue threshold.
My conclusion is that the business objectives of app developers and the app stores are not the same. Hoping and complaining will not change this. Even when the focus is on revenue, there is still a misalignment. For example, consider this assessment of where Google's priority lies:
The app stores, while a tremendously useful distribution platform for marketing and sales, must not be relied upon to assist developers to create a strong business plan. They aren't hostile, it's just that they simply don't care enough. Hoping and praying for new features and resolution of longstanding problems is not acceptable behaviour for those serious about building a business. One way or another, each mobile app entrepreneur (as distinguished from a mobile app developer) must build a business plan that does not have this dependency. The app stores are a tool, not business partners who have a shared financial interest in your success. Write your own ticket by removing this high-risk dependency from your business plan.
A large majority of apps on iPhone (85,000+) and Android (10,000+) are either free or are not generating enough revenue to pay the developer's home utility bills, let alone achieving profitability. The odds for success are much better than lottery tickets, but there is a lot more to being successful than coding like mad for a few weeks and then praying for the money to come rolling in. Tapping into the consumer market is particularly fraught with peril since it is so very hard to predict how the market will respond. This is anathema to developing a workable business plan unless there is already some market success for a core part of the planned functionality from which revenue extrapolation can be reliably attempted. Going after the enterprise market is safer since the prospective customers will buy if they have a good business case for a product, and that determines the app's revenue prospects. Setting yourself up so that you're subject to the whims of consumer tastes of the moment will not endear you to investors.
What we need to consider is just what the business model actually entails when selling mobile apps, regardless of the target market. The majority of mobile apps are distributed through app stores run by Apple, Google, Microsoft, RIM and Palm, for apps that are designed for their own brands of mobile OS platforms. There are third-party stores and possibilities for carrier stores (although Verizon, in their announcement this week, surprisingly to some, claimed their Android phones will rely on Google Android Market), but we can discount these parallel distribution channels for the present. Because of developers' reliance on these app stores for marketing, distribution and payment processing, for which they pay a price (typically 30% of sales revenue, split between the app store operator and the carrier in most cases), it is important to assess whether this "partnership" is to the benefit of both parties.
For the platform vendor (which is the same as the device vendor in the cases of Apple, RIM and Palm), there are two key benefits of an app store:
- Increase the value proposition to users by making available a large pool of applications; and,
- A revenue stream that is additive to device sales or OS licensing.
With regard to the second point -- revenue -- the dynamics are far more interesting, and telling. I have seen a number of contradictory stories of how much money Apple is deriving for their app store (either very large or very small), and how much of that is shared with carriers. Even if the true number is toward the high end, the reported figures still pale in comparison to the revenue derived from iPhone and iPod. That is, it's a nice pile of money but not enough to get too excited about for a corporation the size of Apple. In the case of Google, the amount must be far lower since we know that the majority of the 30% fee goes to the carriers, and there is much lower success rate of paid apps on the Android Market. (Google, as I'll discuss, has little incentive to do better.)
Putting these facts and projections together, I conclude that the primary objective is the first point: increasing the device value proposition by providing users with a large catalogue of applications, both free and paid. The problem for app developers is that Apple and Google have already achieved this objective, and therefore have little incentive to make any further attempts to help app developers to succeed. App developers would obviously prefer that their so-called partners' primary objective were to maximize app revenue, but that is not the present situation. We should expect the same to hold true for RIM, Microsoft and Palm when or if they pass the same app catalogue threshold.
My conclusion is that the business objectives of app developers and the app stores are not the same. Hoping and complaining will not change this. Even when the focus is on revenue, there is still a misalignment. For example, consider this assessment of where Google's priority lies:
"Google views mobile phones as an important way to extend the growth of its online advertising business, as people around the world increasingly access the Internet from smartphones.Of course, Google's advertising revenue objective is not restricted to Android, but to all mobile platforms.
Analysts believe that mobile phone ads have the potential to be more lucrative than Internet ads viewed on PCs, given the that mobile ads can take advantage of an individual's physical location to better target a marketing pitch."
The app stores, while a tremendously useful distribution platform for marketing and sales, must not be relied upon to assist developers to create a strong business plan. They aren't hostile, it's just that they simply don't care enough. Hoping and praying for new features and resolution of longstanding problems is not acceptable behaviour for those serious about building a business. One way or another, each mobile app entrepreneur (as distinguished from a mobile app developer) must build a business plan that does not have this dependency. The app stores are a tool, not business partners who have a shared financial interest in your success. Write your own ticket by removing this high-risk dependency from your business plan.
Monday, October 5, 2009
Selling Mobile Apps: Friction
Giving away free stuff on the internet is easy. Persuading people to pay is difficult; much more difficult. What's a mobile app developer to do? Everyone eventually needs to get paid since, after all, that's how we all make a living. If app development is a hobby, it only means that the developer has an unrelated source of income.
Let's look at sales friction a little more closely in the context of Android, Google's open source OS for mobile and other devices. Further, let's assume we've gotten past the willingness-to-buy challenge for a particular user. Now all we need to do is make the money flow from the user to the developer, via one or more intermediaries such as payment processors. As you can see, I am focusing on one small aspect of a large problem, but one that I believe is worthy of attention.
For the user, the first step is to decide from which account the monies will be paid. The obvious choices for the majority of users would be:
If you are a mobile app developer, you most often use the device or OS provider's own market to sell those apps. This is often not necessary, as Android has shown, although it is the norm; most device vendors make it very difficult to sell apps (even free ones) any other way. Let's look at a couple of cases of mobile app store payment processes.
Apple's iTunes store was established long before the iPhone came onto the market. Users signed up to iTunes by the millions so they could purchase music for their iPods. The draw of the iPod and digital song downloading was so strong that friction in the account creation process was easily overcome. Had Apple messed this up, iTunes would not have had the same degree of success. When apps became available for the iPhone and more recent iPod devices, most users already had an iTunes payment account. The friction to purchase mobile apps was low, leading to stunning success for Apple and many app developers.
Google set up the Android Market with Google Checkout as the sole payment processor. Google Checkout has a small market share and suffers from a variety of problems, including many that are unrelated to Android Market. Few Android phone users were therefore likely to have had a pre-existing Google Checkout account. The friction when a user attempts to purchase their first Android app is high: sign up with a new payment processor, one that they may have heard bad things about, reading and agreeing to a long list of terms and conditions, handing over a credit card (but not a debit card), and then (at long last!) make the purchase. At this point you may be wondering just where the friction is; the extra steps involved may not seem all that big, or at least you want to believe they're not big if you're an app developer.
Unfortunately, these are big steps. There is a reason the carriers, Amazon and other mass market merchandisers work so hard to simplify the payment process: every bit of friction has a large impact on revenue -- this is not only true for payment, it is also true for operation of a product or service. This is one reason why so many large companies invest a lot on marketing and sales, and less (sometimes much less) on post-sales customer service: they know where to invest to maximize their return on capital.
Talking to some of these large corporations -- as I have throughout my career -- helps to understand what they have learned through hard experience. A common refrain is that for every button press or mouse click you lose a percentage of the market. Sometimes it is a small percentage, but more often it is surprisingly large. Add these things up and the potential market penetration of a service (or app) can become very poor. This is the crucible they throw product ideas into to determine if they have a chance for mass market appeal. If not, it gets categorized as a niche product, and their interest in pursuing the idea drops dramatically. That is the impact of friction.
On this basis, the friction to sell apps on the Android Market is moderately high, as many app developers have unhappily discovered. Unfortunately the currently-available alternatives have even higher friction. Other app stores like AndAppStore and SlideMe have the added friction of requiring user awareness and willingness to set up an account at stores with an unknown brand. It doesn't have to be that way, but that is the present situation. The friction can be even higher for a developer-operated payment system, with even lower awareness, lower trust, and the need to pull out a credit card for every app purchase. Again, there are ways to reduce the friction, but it can never equal that of a major brand app store that comes packaged with the consumer device.
With what I know of the situation with Google Checkout, I do not expect that its current problems will be resolved anytime soon. For app developers, the better bet may be carrier-operated app stores. My ideal scenario would be a carrier market application bundled with the phone to connect with the carrier-operated app store, with an ability to "see through" to other app store partners, including Android Market, and integrated with the carrier's own billing system. That should sufficiently lower the sales friction to a level comparable to iTunes, which should give priced Android apps a real shot at commercial success. At that point Google could pretty much stop investing in its own app store -- in which it is showing a great reluctance to improve. Google could then solely focus on the Android core platform, where their talents can be better applied.
Let's look at sales friction a little more closely in the context of Android, Google's open source OS for mobile and other devices. Further, let's assume we've gotten past the willingness-to-buy challenge for a particular user. Now all we need to do is make the money flow from the user to the developer, via one or more intermediaries such as payment processors. As you can see, I am focusing on one small aspect of a large problem, but one that I believe is worthy of attention.
For the user, the first step is to decide from which account the monies will be paid. The obvious choices for the majority of users would be:
- Credit card
- Debit card
- Mobile service provider account
- PayPal, Google Checkout, iTunes or other 3rd party payment processor account
If you are a mobile app developer, you most often use the device or OS provider's own market to sell those apps. This is often not necessary, as Android has shown, although it is the norm; most device vendors make it very difficult to sell apps (even free ones) any other way. Let's look at a couple of cases of mobile app store payment processes.
Apple's iTunes store was established long before the iPhone came onto the market. Users signed up to iTunes by the millions so they could purchase music for their iPods. The draw of the iPod and digital song downloading was so strong that friction in the account creation process was easily overcome. Had Apple messed this up, iTunes would not have had the same degree of success. When apps became available for the iPhone and more recent iPod devices, most users already had an iTunes payment account. The friction to purchase mobile apps was low, leading to stunning success for Apple and many app developers.
Google set up the Android Market with Google Checkout as the sole payment processor. Google Checkout has a small market share and suffers from a variety of problems, including many that are unrelated to Android Market. Few Android phone users were therefore likely to have had a pre-existing Google Checkout account. The friction when a user attempts to purchase their first Android app is high: sign up with a new payment processor, one that they may have heard bad things about, reading and agreeing to a long list of terms and conditions, handing over a credit card (but not a debit card), and then (at long last!) make the purchase. At this point you may be wondering just where the friction is; the extra steps involved may not seem all that big, or at least you want to believe they're not big if you're an app developer.
Unfortunately, these are big steps. There is a reason the carriers, Amazon and other mass market merchandisers work so hard to simplify the payment process: every bit of friction has a large impact on revenue -- this is not only true for payment, it is also true for operation of a product or service. This is one reason why so many large companies invest a lot on marketing and sales, and less (sometimes much less) on post-sales customer service: they know where to invest to maximize their return on capital.
Talking to some of these large corporations -- as I have throughout my career -- helps to understand what they have learned through hard experience. A common refrain is that for every button press or mouse click you lose a percentage of the market. Sometimes it is a small percentage, but more often it is surprisingly large. Add these things up and the potential market penetration of a service (or app) can become very poor. This is the crucible they throw product ideas into to determine if they have a chance for mass market appeal. If not, it gets categorized as a niche product, and their interest in pursuing the idea drops dramatically. That is the impact of friction.
On this basis, the friction to sell apps on the Android Market is moderately high, as many app developers have unhappily discovered. Unfortunately the currently-available alternatives have even higher friction. Other app stores like AndAppStore and SlideMe have the added friction of requiring user awareness and willingness to set up an account at stores with an unknown brand. It doesn't have to be that way, but that is the present situation. The friction can be even higher for a developer-operated payment system, with even lower awareness, lower trust, and the need to pull out a credit card for every app purchase. Again, there are ways to reduce the friction, but it can never equal that of a major brand app store that comes packaged with the consumer device.
With what I know of the situation with Google Checkout, I do not expect that its current problems will be resolved anytime soon. For app developers, the better bet may be carrier-operated app stores. My ideal scenario would be a carrier market application bundled with the phone to connect with the carrier-operated app store, with an ability to "see through" to other app store partners, including Android Market, and integrated with the carrier's own billing system. That should sufficiently lower the sales friction to a level comparable to iTunes, which should give priced Android apps a real shot at commercial success. At that point Google could pretty much stop investing in its own app store -- in which it is showing a great reluctance to improve. Google could then solely focus on the Android core platform, where their talents can be better applied.
Subscribe to:
Posts (Atom)
Posts
