Goodbye to Google Optimize

This week we have had nothing but scares with Google. Optimize is over. Google has communicated that, despite the time that the tool has been with us, the product “does not have many of the features and services that our customers need for experimentation tests. We have therefore decided to invest in solutions that are more effective for our users”. So, as of 30 September this year, Optimize will cease to exist.

Google has already implemented the move with links for exporting test data and raw data from the Analytics API. Exporting data from a tool is a bit like moving house. In the long term it’s a good opportunity to clean up and get rid of things you don’t want – some people have gotten rid of couples like that – but in the short term it always leaves a feeling of nostalgia. Tempus fugit.

The link to Google’s note about the end of Google Optimize here.

We are not (so) surprised either

The rumor had been making itself heard for days. Possibly the strongest signal came from Google, which closed a deal with Optimizely, but there is no reference to it in Google’s press release. So, although some people interpret the end of Optimize as a move similar to the “replacement” of Google Data Studio by Looker, right now it seems that this is not exactly the situation.

Google Optimize will cease to exist as of 30 September and Google is not proposing an alternative. Optimizely has a partnership agreement with Google, but it remains a separate project and there seem to be no plans to integrate it into the suite. In fact, Google’s comment is that it will develop integrated A/B functionality within GA4. Optimizely remains there, as just another tool. Their collaboration with Google will probably favor integration with other Google tools and I don’t doubt that they will be at the forefront of adapting to GA4 changes. But it doesn’t look like there will be any integration, not even at a naming level, as has been the case with Looker.

The end of Optimize =! the end of optimism2

It is almost certain – to say the least – that the end of Optimize is not a spur-of-the-moment decision and it is very difficult to know to what extent the decision may be related to other news that Google has made during the week and to the general context in which it has moved in recent months.

Because this week Google announced the biggest job cuts in its history: 12,000 employees worldwide. A few weeks ago, after Facebook made its own job cuts, there was talk that Google had prepared lists of employees to rate them on their performance. It was said that the aim was to save money on productivity bonuses and that this would be Google’s strategy to reduce salary costs without actually firing people. Well, the layoffs have come. So Google has not escaped what already seems to be the end of the uninterrupted growth of the big tech companies. Amazon, Facebook, Twitter (as if they didn’t have enough) and now Google already know that growth cannot be infinite and are implementing cost-control measures. And if the economy, as they say, is a state of mind, it remains to be seen how this sudden melancholy that has hit the tech giants affects the general economy.

Is Optimize a victim of this new scenario? Maybe not entirely. It seems more like an epiphenomenon. Optimize never really established itself as a tool as commonplace in the digital marketing world as other tools in the suite. When it came on the market, it did so under the pretense of doing for optimisation and user testing what Analytics had done for analytics in general: democratization that would take A/B testing out of the dark workshops of technical departments and into the bright rooms of marketing offices. But it hasn’t. Optimize never got past the technical barrier. There is an unwritten law, but one that has so far proven to be as solid as gravity: marketing departments are not willing to accept a single line of code in their brightly lit offices. Not a single line of code. Zero. If a tool means that a marketing expert has to deal with even one measly HTML tag, it will immediately be banished to the backwaters of “the IT guys”.

The graph below shows the evolution of the searches “Google Data Studio” and “Google Optimize” over the years. These are Trends data, so we don’t see absolute volumes there, but it is clear which of the two tools has been improving and which has remained stagnant for years.

And the truth is that Optimize is a good tool. For many companies, it was an ideal tool to get into the world of A/B testing and optimization. Of course, it had its limitations, but for many companies it was sufficient. However, it had a couple of problems that I think have definitely hindered its implementation.

Apart from falling into the innocent assumption that someone from a marketing department could cope with looking at a script without the support of a qualified exorcist, the truth is that the tool was ultimately useful for a very specific – perhaps too specific – range of users. Below a certain level of digital development or resources, companies were not able to use the tool productively. A company that pays little attention to its own digital analytics is unlikely to invest the resources and patience that an optimization plan needs.

At the other extreme, a sufficiently developed company with ambitious optimization goals and a sophisticated UX plan will soon find the limits of the tool. You could say that GA or Data Studio Looker Studio has the same problem, but here we come back to the fact that those are tools that can be used with zero technical knowledge and without relying at all – or almost – on development teams. In the case of optimisize, the dependence on development teams was much greater, especially in the areas of websites with certain functionality, which are precisely those that can benefit the most from optimization tests.

The end of Optimize comes when very important changes are taking place around the GA suite, which will inevitably produce very important changes in the world of analytics in general and in the entire digital sector. For starters, the move to GA4 is almost imminent and the general feeling is that there is still much, much homework to be done. Google has worked hard over the last year to adapt the tool and make the transition smoother for UA migrants. But the truth remains that the change from UA to GA4 is a paradigm shift, that many companies are not at all ready for the transition and that there are still those who think it is like moving from Windows 95 to Windows 98.

On the other hand, Google has sounded the alarm. The emergence of ChatGPT and, in general, the release of several AI-based tools seems to have caught Google on the wrong foot, which is curious considering that Google has been at the forefront of the most spectacular advances in AI over the last few years, at least in terms of neural networks. It is quite possible that this feeling has more to do with a commercial and communication issue than with a technological breakthrough. But the general feeling seems to be that Google has been overtaken by the right and that its next move (it has just announced a batch of AI-based tools for 2023) is more of a reaction than a decision, while GA4’s exit seems to be more related to legal harassment than to Google having a hard time abandoning a product (UA) that with its good and bad things has achieved a very powerful leadership in its category.

Miguel Carreira López
Miguel Carreira López

I have been working in web analytics and SEO for ten years. I believe that there is nothing that can be said about the traffic of a website that can't be better explained with a graph. I work mostly with Google tools (GSC, Looker, Analytics) but there is life beyond that. In my spare time I write about books at