Comparing Google Optimize alternatives

Optimize was free, but for a reason

Google Optimize was late to the experimentation party but quickly became the market-leader in terms of volume of customers, scooping up almost 50% market share. This was primarily based on the fact that it was free, could be deployed via Google Tag Manager (GTM) and had a native integration with Google Analytics.

However, with the tool being free for 90%+ customers, there are notable limitations to the product. There are simple limitations in terms of how many concurrent tests can be run (five), how many goals can be added to an experiment (three) and even how many characters can be used in the code editor (10,240, if you were wondering). All of these are speedbumps for organisations looking to do more – whether that’s simply increasing the number of basic tests being run, or increase the sophistication of those tests.

Likewise, there were limitations in terms of not being able to upload images directly into variations – they instead had to be stored in your CMS or online to be used. And of course, whilst the integrations with other Google products were strong, Optimize failed to build out that ecosystem, so if you used other tools, maybe it wasn’t your best friend.

Either way, like it or not, Google has decided that one sunset this year wasn’t enough (RIP UA) and Google Optimize will no longer be available after September 30th. 

So, what to look for in a new platform?

The first thing to say is that this is not a comparison of vendors, apples-to-apples. In our experience, it’s almost impossible to compare, since different vendors have such different approaches. Whilst Web Trends Optimize keeps things simple in having no tiers, just pricing on volume of website traffic, VWO is far more granular, but complicated, in what you can buy, having six different products with 14 different plans.

But what we can highlight is what you should be considering, and asking of vendors, when you’re speaking to them:

1.     Cost – this will always be the biggest factor, or at least play a very significant role. However, as per the above, the old adage is as true as ever. You get what you pay for. VWO offers a free package for sites with up to 50,000 users but the majority of other tools will charge something. Is this good? If you’re suddenly looking for budget for a tool that was previously free…then no. But as an industry, and as an advocate of experimentation in your business, it does mean there’s an opportunity to really get senior investment in your program by investing in a tool to take you to the next level.

2.     Targeting options – what sort of tests do you want to run? If you’ll be running tests to everyone then this won’t be a factor. But if you want to be more discreet, running tests against certain type of users, coming from certain types of campaigns or even at certain times of the day, you’ll want to pay attention to the packages, and ask the right questions during your demos. We’ve consistently found that “symmetrical messaging” – where you tailor a landing page based on the channel someone has come there from – is a fantastic way to make experiences feel more seamless and drive conversion rate. But it’s not available on the basic package of all vendors so make sure to check.

3.     Native integrations – plenty of platforms can potentially build integrations, but making sure that the vendor you’re speaking with will integrate with the other tools you use is really important. Many will also talk about how easy it is to build integrations  – and whilst that might be true sometimes, in other cases you’re going to want to be sure the integration works so an existing connection is far more compelling. Likewise, listing a platform logo is one thing – the nature of that integration is another. Sometimes it’s flicking a toggle button in the interface and you’re flying; other times it’s grabbing IDs from one platform, sharing with another, resolving somewhere central and then reuploading some IDs into the testing platform. Perfectly doable but not exactly smooth!

4.     Plug-and-play widgets – this is a bit of a catch-all but many of the tools offer “accelerators” to your testing program. These might be out of the box ways to configure social proof, or recommendations; or they might simply be reusable assets that allow you to deploy pop-ups or USP bars more quickly than if you had to code yourself from scratch. Check the tier you need to access these features and also find out how self-serve these really are. In our experience, sometimes these are very native, and in others, it’s a bit of a plea to the vendor’s services team.

5.     Deployments – something else which has a different name across almost all vendors but this speaks to the ability to run a winning variation to 100% of traffic whilst your developers build out the full version on the live site. It’s also linked to the number of concurrent tests you can run, in some platforms, as that’s effectively what you’re doing (except 100% of traffic sees the version you want them to). With some vendors this capability is baked into all packages whilst for others, it’s tied to a particular tier or available as a wholly separate add-on.

6.     Support you’ll need – how much help will you need to onboard the new tool? Is the deployment complex? Will many people need to be trained? Is there an additional cost for this!? These are the sorts of questions to be asking as you go through your vendor assessment. Some tools will make you pay for onboarding support whether you want it or not; others are more flexible or let you choose a partner to offer the services whilst others will let you dive straight in.

Think about a realistic roadmap

There’s lots to consider when choosing a new tool, even more so if you’re moving from the free version of Google Optimize. But the critical thing is to think about your experimentation program as it is now, the tools you currently use and about the likely types of tests you’ll want to deploy in the next 12-18 months. This will give you a good shortlist of features to focus on, whilst being realistic as to what you’ll use. Be sure to check if it’s possible to bolt-on certain features too, if you find you need them in eight or nine months.

If you want to chat through options, or even need support running a vendor comparison, give us a shout!