Deconstructing Google's Trust Problem

Right after it announced the execution of Google Reader, Google released a new product — really a service — called Google Keep. It basically is an Evernote, note keeping app on the Google ecosystem. This prompted many bloggers including The Atlantic’s James Fallows and Ezra Klein to announce their reluctance to invest in using products like these as they cannot trust Google not to kill them in the future. Here is Fallows:

Google now has a clear enough track record of trying out, and then canceling, “interesting” new software that I have no idea how long Keep will be around. When Google launched its Google Health service five years ago, it had an allure like Keep’s: here is the one place you could store your prescription info, test readings, immunizations, and so on and know that you could get at them. That’s how I used it — until Google cancelled this “experiment” last year. Same with Google Reader, and all the other products in the Google Graveyard that Slate produced last week.

While I too have become worried about Google’s services, I think people need to get a grip.

First of all, the demise of Reader is not nearly as bad as it could have been. As many have pointed out, subscriptions and feeds are easily exported to other services and, in any case, Google based the whole thing on an open standard, RSS. If it had all been proprietary, we could really have been in trouble. Of course, in that case, we may not have invested in using Reader in the first place.

Second, it is interesting to see that many commentators are engaging in strategic reasoning with higher-order beliefs. The low-order belief would be, Google killed products — Google bad — never trust Google again. A higher-order belief is Google killed Reader — Reader didn’t support ad revenues — don’t trust Google to keep products it doesn’t earn money on. Or, similarly, Google killed Reader — Reader didn’t assist in search — don’t trust Google to keep products that don’t support its core search business.

Suffice it to say, lower order belief reasoning is naive and likely self-destructive. But higher order belief reasoning is a mess. Think about it. Every time Google releases a new product, we have to engage in a strategic analysis of its likely longevity. What is more, we all have to do it and we all pretty much have to engage in the same analysis in order to actually tease out a solid prediction. (Actually, maybe not the same but you see my point). No one really knows what the heck Google is really doing, including Google itself with its desire to experiment. Every time something new is released we who profess some experience in higher order strategic reasoning can barely agree on anything. What is the rest of the population supposed to do? When it comes down to it, we quickly get back to: “This product works, is useful and is cool, maybe I should use it.”

So I disagree with suggestions that we should expect people to conduct strategic analysis to adopt a product. This includes not just products by Google but pretty much anyone. Google isn’t unique in offering services and this is just an issue with any product that you might put into your daily habits. It makes the sense of loss greater when it goes.

I think what is frustrating about Google here is that they are not using their incumbency advantage to help us all along. If some start-up launches a great product but in the end can’t make money off of it and it closes, there is sadness. This can also make us nervous. For instance, in the early Amazon days where it was critical for getting books to Australia, a colleague of mine wrote to them begging them to increase their shipping rates. She didn’t want them going out of business just because they were trying to build market share. These concerns are no longer there with Amazon.

But Google is treating these new products as if they are in a start-up mode. “Oh dear,” they say, “usage has declined. Well, a start-up based only on that product would shut down and so we will be start-up like and do the same.” But they are not a start-up. That gives them a credibility advantage. If they are going to shut down a product, because they are going to be still around they can tell us more about why they are doing it. Reader is going and there is no substitute offered by Google. So why not actually publish the usage statistics. At the moment, everyone thinks that this is a mild decline but the product was fundamentally sound. Maybe it wasn’t. Maybe no one was really using Google or the decline was dramatic or the costs of providing it where just too great.

My point is this. Google’s trust problem is not that it will kill products that don’t serve its strategic interest — even if we could distill it. It’s trust problem is that it may kill products on a whim. Everything that it communicated with regard to Reader suggested more of a whim than real strategy. Google can do better than that. It should be able to kill poorly performing products without suffering a trust issue. That is what incumbents do. Maybe in this regard, it should take a leaf from Apple, Amazon and Microsoft. When they kill products, people usually understand or even better, don’t really notice.