Optimizely One vs. Twilio Segment

It is well suited for heavy-duty A/B testing where management would like to see and quantitatively determine the effect of a change. It is not so efficient to try for a single page simple form, except when the form is part of a larger workflow. The security model is not very well understood, including RBAC and protection against injection attacks.

Incentivized

Best suited: - Merging emails coming from: Facebook leads forms, Unbounce or landing pages forms, Google forms, any other kind of lead generation tool and bundling all that information together for a single user "profile". - Passing events generated in multiple applications by the same user (product selected in web, product discarded in cart, etc) and delivering those events into other applications (like a CRM) Less appropriate: - Reading/updating data directly from segment from a frontend application

Incentivized

• Provides quicker release cycle for the way the product cards should be displayed.
• Provides feedback for new product introduction and how it'll be perceived.
• Allows rapid blackout from user experience changes that alters traffic flow.

Incentivized

• Multi-platform. Segment has easy integrations in many different web, backend, and app platforms/frameworks. We use the Segment SDK in Android and iOS as well as our node.js backend.
• Segment is fairly affordable for early-stage companies that are trying out different analytics software. The "developer" plan is free and is suitable for most companies with products that have a small user base.
• The UI is great! It is extremely intuitive and easy-to-learn, and this made it take very little time to integrate this software into our analytics and marketing workflows.

Incentivized

• Can use better interface for managing test, including scheduling and notification.
• Feels too heavy for simple projects, though this isn't a factor once past the PoC stage.
• Notification on various aspects can be made more powerful and granular if such a need arise.

Incentivized

• More and richer sources. For example, MailChimp is a source but the data you get from MailChimp is quite limited. I ended up writing my own scripts to take better advantage of MailChimp's API because Segment's integration was lacking.
• Better examples on how to set up event tracking. Pageview tracking is easy enough, but it would be nice if they had a sample app and corresponding code for it and showed you, via Git commits, how to add various kinds of events.

Incentivized

Over the period it took us to set up, we kept going back to their enablement team to help us with the setup, and they were always ready and were very helpful in the entire process. Even with their documentation, they took the time out to help us work through the process. We've never had a message/email unanswered for more than an hour on working days.

Incentivized

There is a big difference between the two: Google Optimize uses Bayesian analysis while Optimizely uses Frequenting. There is a risk of counting multiple visits. Google Optimize data isn't available instantly (if I remember correctly). Optimizely's analyses dashboard is a lot richer and offers a better experience, though it may get intimidating.

Incentivized

We chose Twilio Segment for the good API integration and node resources, I would use Ontraport again, particularly if I didn't have the requirements for API and development/platform integration. Certainly the set up and management is easy and seamless with both the API and the user interface to use depending on circumstances and requirements.

Incentivized

• Almost immediate payoff in terms of the benefit.
• Easy to justify the TTM benefits to upper management.
• Still somewhat hard to justify ROI, though this is not my expertise.

Incentivized

• Segment has enabled us to get a full view of our front end activity, join it to our back-end activity, and get full visibility into our funnels and user activity.
• Segment lets us send events to ad tools with a full audit trail so all the numbers line up.
• Segment also brings data from other sources into our data warehouse, saving our data engineering time from building commodity connectors.