The power of a universal login — a history lesson from my time at Yahoo

First published on the miggle blog 25/4/17

I worked at Yahoo between 1999 and 2006. I’d say for the first two years we had a pretty much unparalleled opportunity. For the next two years it was probably on the verge of still being able to deliver on that, despite some existential challenges, and then for my last three years there I would say that it started down the long road of utterly losing its way. (Read about Brad Garlinghouse’s Peanut Butter Manifesto for a great commentary on that from 2006)

In this post I write about one of the key mistakes I think we made, because I think it carries lessons which are still very much relevant for today’s modern web. Who knows? This may be the start of an exciting new series of stories…

Anyway, for now, this story explains why it was critically important for Yahoo to maintain a single universal login that allowed users to access all of its services and how its failure to maintain this was a mistake. It should hopefully provide some historical context to anyone who is considering the challenges and considerations around managing a single login to allow access to all of your services.

Yahoo was always an advertiser-funded network which primarily sold audience. For that it needed a large and engaged user base.

Alongside having the right content, one means of trying to get effective engagement, as well as building the value of the audience to advertisers, was to encourage those users to register to use features that Yahoo ran within individual products like Sports (e.g. play fantasy football) or Finance (e.g. manage stock portfolios). These in themselves opened up opportunities for premium revenue streams, where users would pay for more advanced functionality within those features. The thinking was the more unregistered users you can get into the top of the funnel, the more of those would convert to registered users, the more of whom would convert to premium users. And of course apart from the funnel growing in size there were various strategies you could adopt to try and improve the rate of conversion between unregistered, registered and premium.

There were also transactional revenues to be made; a percentage on sales generated within commerce products like Shopping, Auctions or Travel. It’s fairly evident where having registered users for those kind of products would help Yahoo maximise sales.

And then of course Yahoo already had a whole range of online tools for which registered usage was essential, Yahoo Mail being by far the largest driver of registrations. Of course a product like Mail had its own significant premium revenue opportunities too.

Pretty much every product had a registered user strategy and it was clear that increasing the number of registered users would have a direct positive impact on advertiser, premium and transactional revenues.

Imagine the awesome power and opportunity of being able to access all of those products with just one username and password: Your universal, network wide Yahoo ID!

There was one other very lucrative revenue stream. Revenues could be generated by forming effective partnerships. Before Yahoo went into or reviewed any viable product business we’d ask ourselves this question: do we build it, buy it or partner with it? This made a lot of sense as a business, particularly in Europe where we were headcount limited. In an ideal world it would have been great to build every viable product, but there just wasn’t the scope to do that — and of course in many areas there were already very well established players. Buying these players was often an option; we certainly had a significant enough war chest to enable us to do so. But Yahoo never acquired or integrated businesses particularly well. In any case the easiest deals to get going were partnerships, because they could be driven at a less senior level than acquisitions or new builds.

A very good example of this was classified products in UK, primarily Jobs, Cars and Property. Internally we had tried to build these businesses, but there just weren’t the resources behind them to deliver a product that could match that of our competitors. Consequently audience numbers were very low and so was revenue. Partnering made perfect sense. Yahoo would get quality content and functionality out of the deal, which it could more effectively monetise with banner advertising, which was then split on a revenue-share basis.

The thorny issue was always around who owned the customer. So when a user registered to use advanced features of one of the classified products, were they a customer of Yahoo, a customer of the partner or both? In all instances that I remember, Yahoo far too easily traded ownership of the customer to keep a higher percentage of the advertising revenue share.

But there was another more important reason why ownership of the user was also traded or lost. For a partner to be able to integrate its user registration technology with Yahoo’s universal ID required some level of technical development. That integration cost money because it took time and both factors would inevitably hurt the bottom line of the deal for Yahoo. Even if it was a go-er, getting a foothold on the partner’s technology roadmap was often difficult. And in any case where was the partner’s incentive? Why have that integration headache to enable Yahoo to have some registered ownership of the customer, when for you as a niche player the focus of building your user base was more important than hitting a few short term revenue goals?

So what happened? In the end we ended up with too many situations whereby a universal Yahoo ID didn’t apply to all products. The whole point of a single login was lost and swapped for a fudge where, under Yahoo branding, a user signed up and signed in to our partner instead. It was confusing for users who did not understand why changing a password in one area (like Jobs) didn’t seem to roll over everywhere (like Finance). And for those who didn’t understand why some products required your username to have your domain extension, some didn’t and on others, your username was completely different. That’s before we even get into the various different user experiences for managing and registering accounts. Trading the sanctity of the universal ID was utter lunacy. Can you imagine in 2017 Google, Ebay, Amazon, Apple or Facebook doing the same thing? No. Which is why they are still here in force.

Ultimately the business reasons behind this were that sales and biz dev functions were primarily rewarded on a quarterly basis depending on whether they hit their numbers or not. And if you didn’t hit your numbers for more than a couple quarters in a row there was little point you staying around. There was very little incentive from a revenue generating perspective to do anything that protected the integrity of the features that allowed users to universally manage their engagement with products. Of course the longer term effects of this over time were that it eroded the quality and quantity of the audience, which of course had a direct impact on the ability to make advertising revenues in any case. And of course Yahoo built up competitors well past the point where we could compete with them or acquire them.

This is all one very good historical lesson as to why having a universal ID to access your services is absolutely essential for a whole number of reasons. Failing to do so was one of the key ways in which Yahoo started to squander its opportunity.

Digital decisions are never a walk in the park, so let me help you find the right way through the technical landscape.