Sunset of Google Optimize September 30th 2023 - What you need to do and what does this mean to CRO?
Google has announced that Google Optimize will sunset on September 30th, 2023. This, according to Google means that “Your experiments and personalizations can continue to run until that date. Any experiments and personalizations still active on that date will end.” I will not be covering your plans or what happens to your licence. For that information see the Google Help Center.
A lot of people seem to think it will be “relaunched” in a new way and integrated with GA4. Some think that it will take a few years. Some think that they are looking into new technology now but it will take some time. Since GA4 might not be where we want it to be yet - do we really think they will have a solution in place in time for the sunset? Perhaps not. And if you ask me, I wouldn’t take the risk.
What do I need to do now?
If I had a testing program ongoing with Optimize, I would ask myself; would I take that risk? And if the answer is no, what does the timeline look like? We need to calculate how much time it would actually take - switching towards a new platform.
Having switched AB-testing tools quite some times before I would start by setting a decision date. A date for the last day possible for your organisation to make that decision; to move or not to move from Google Optimize.
Let’s say you have a CRO program running, with about 10 tests being live simultaneously. These probably run within the average 2-4 weeks. But then you have to take into consideration how long does that test take in its full? So from the moment you decide to move forward with a hypothesis until the moment it’s analysed and done. You might want to stretch that to a 2 month period. Just to be safe. So if you’d want to move over organically - you need to have your setup (including AA-tests done) 2 months before September 30, 2023: Which calculates to your date; July 31st. After this date - you should organically start moving all your new experiments to your new platform.
Of course there are a couple of steps to do before that. Like choosing the right platform for you, going through legal and GDPR documentation, actually signing that new tool. Perhaps this also includes you needing to talk about the budget in the organisation - if you did use Optimize for free that is. Well, after you have signed you need to get the snippet implemented - and of course integrations to your analytics tool, but also make sure your new AB-testing tool is gathering the data you want to use. You of course then need to check your data quality and run your AA-tests.
Here’s that checklist for you:
RFP and requirements for a new tool
Set the requirements for you testing program (but if you used Google Optimize before you should be good enough with most tools)
Budget allocation
Legal and GDPR documentation
Implementation and integration to analytics tool
Data quality check - does your data in your new tool differ from your analytics tool? What data will you be using for which metrics?
Run AA-tests. This helps you learn your new tool, but is also essential for quality check of the data and level of significance
Ask stupid questions. Every AB-testing platform works differently, has different icons and shows your results in different ways. Ask your new provider about what it means, so that you can take the best decisions
So the question is, how long will this procedure take for you in your organisation? Take that time and subtract it from July 31st. This is your true sunset date.
Additional note on the Data: Merkle also states that you will be able to use your Google Analytics data integrated with your experiment platform, but this is also then up to the providers to collect this data. And this is great - should perhaps have happened a while ago - but it is also a new way of collecting data to your AB-testing tool. That we also need to take time to and understand. When doing this switch you might need down time for all your experiments. Or could you work with multiple data sources for different experiments at the same time? I guess this is a question for our providers - and a question for you to hold on to.
Market shares and AB-testing tools. What will happen with CRO in a post-free-AB-testing-tool-world?
Future Market Insights claims that the AB-testing Software Market is estimated to grow 11.6% between 2022 - 2023. And this change in the landscape might give that an extra push earlier than they thought.
To understand what will happen after the Google Optimize sunset - we first need to understand what the market looks like today. When investigating market share it quickly became clear that most research does not even include Google Optimize in their reports. This might be because Google doesn’t release these numbers, although the actual reason doesn’t matter. What matters however is that the reports are somewhat useless. Because these reports leave out a substantial part of the AB-testing world.
According to Convert Market Share who does include Optimize, they stand for a market share of about 61%. BUT they also state that only 2,4% of these actually have active AB-tests. I’m guessing that number is higher for tools like Optimizely and VWO - where you actually pay for the service. But of course, these tools will also have some under performing users. However this report does not include Adobe Target - who according to Datanyze also have a significant market share.
The number of companies using Google Optimize according to Datanyze is 9382 (see picture below), compared to Convert’s analysis of Optimize running on 6,759 active domains. I think we’re getting somewhat close to the truth here.
But to be sure I also asked ChatGPT (mostly for fun), who also stated the same conclusion as I have: that Google Optimize, Optimizely and VWO are considered the top players with significant market share.
My guess is that it will be higher if you pay for it. That is - you made a decision to actually invest in this - then you are more likely to use it. That said - if it is true that Optimize has about 60% of the world’s AB-testing people at hand. They are more likely not to utilize their tool in an efficient and active way to provide value. And according to this data gathered - they most probably have about 7% market share - excluding the inactive users.
When sunset hits September 31st, 2023. We will see 2 groups of users: where one of them will move forward with a new testing tool, and the other will hit the pause button on their AB-testing program.
While the group that hits pause for 100% of their users might be large (basically the 93% who don't actually use Optimize + additional). There will be a big group of organisations that will move onto a new platform. The platform chosen will most probably be with a pay plan. As none of the big platforms have had the offer of a free plan. We will however, most likely, see our well known platforms taking actions and steps towards free och much cheaper price plans. Just as VWO announced on January 22nd - a free version of the tool up to 50k users per month. The pricing and cost of running a CRO-program will for sure be the only variable here to why organisations decide to continue their testing program or not. I think we will see a big change in the AB-testing platform landscape this year!
So this means that a lot of formar Optimize-companies will probably not be connected to an AB-testing tool for a while, it doesn’t have to be a bad thing for CRO as a whole. I would argue that we will see CRO in 2023 getting a good growth in maturity. Because if organisations pay for a tool - they are more likely to put effort and time into getting value out of it. So perhaps for us practitioners who just want to create true data driven organisations - this is a great thing!
GOOD LUCK!
CRO-hugs 🚀
Are you unsure about your options or how to move forward? As an independent consultant I can of course help you out. Send me a message, DM on LinkedIn or an email 👇
