Why B2B SaaS needs dedicated feature flagging
If you’re ever used a feature flagging tool for a B2B SaaS product you might have noticed things feel kind of awkward. It mostly works, but even common actions seem slow and clunky. That’s because most feature management solutions are built around users and e-commerce use-cases. Besides generally de-risking deployments, they’re optimized for conducting A/B experiments and checkout funnels.
For engineers in B2B SaaS, that stuff isn’t useful, for multiple reasons:
- Your customers aren’t one-time shoppers. They’re subscribers. Ensuring that your subscribers are happy and retained is your priority.
- You likely don’t have enough volume to conduct A/B experiments with statistical significance. Even if you did, building a SaaS product is about taste. You’re optimizing for overall product UX, not isolated user journeys.
- When it comes to setting up flagging rules, your primary entity of interest is the company who owns the subscription, not the individual users.
- You deeply care about making your customers happy and therefore want as much qualitative feedback from your users as possible.
So, what can be done about it? This post outlines four benefits of feature flagging purpose-built for B2B SaaS and three concrete examples of how Bucket can help.
Let’s dive in.
Increase R&D productivity
B2B SaaS companies invest a lot in R&D. Unlike ecommerce, who typically spend around 15% of revenue on R&D, SaaS companies spend upwards of 23%. In fact, in product-led organizations, this can be as much as 50%. Since A/B experiments are useless in B2B SaaS, you need another way to ensure customers are happy with the outcomes of all that investment in R&D.
Over the past decade, we’ve gotten really good at shipping features. Co-pilots, code reviews, CI/CD and push-to-deploy enable developers to go from idea to merged and deployed code faster than ever. But what happens after we deploy and mark a project as completed in our issue tracking systems? That’s when the customer experiences the feature for the first time.
It’s objectively the most critical part of the feature’s lifecycle, and that’s exactly when we drop attention. It’s why most features fail to make an impact.
To ensure R&D is optimizing for solving customer pain, not just output, we need a way to test beta features in the real world. In B2B, the best way to do that is by getting user feedback as fast as possible.
Using a feature flagging tool that’s built for this workflow enables R&D to ship beta features without friction and to follow through on customer issues whenever needed.
Spend less time bug fixing
The cost of a bug fix isn’t static. When you’ve worked on a feature for months, you have all of the details of the feature loaded in hot memory. But as soon as you move to the next project, those details are gone.
To keep engineering teams moving fast you need to get customer feedback on new feature releases shortly after releasing, so that they can fix any issues quickly as they come in, before it gets lost.
For example, changing a database column type can be done and tested within a few days, if done shortly after release. If not, it becomes a multi-week, high-risk project involving database changes, data migration and customer communication.
With a feature flagging tool that enables a workflow to support this, not only do you spend less time bug fixing and context switching, but you also make your customers happier thanks to the fast resolution times.
Reach Feature-Market Fit faster
There’s a lot of talk about Product-Market Fit (PMF) but not so much talk about Feature-Market Fit (FMF). However, products with PMF need to constantly innovate to stay relevant and ahead of the curve. The way to do that is by shipping new features that land with the folks using them.
But most features aren’t a hit in the first iteration. They’re too rough around the edges, too confusing to use, or something else.
Knowing if new features resonate and get adopted is critical to understanding whether the product is innovating in the right direction. Leading indicators such as early adoption metrics and qualitative feedback are great for predicting feature success in just a matter of weeks.
Being able to quickly determine if a feature has potential or missed the mark is how you move fast as an engineering organization. Feature flags are very useful for turning new features that aren’t a hit back off or to start deprecating them by restricting access only to current users of it.
Skip the Yak Shaving
That awkwardness I mentioned earlier stems from the foundational choices made when building out the feature flagging tool. If the underlying database model isn’t a fit for the B2B use-case, all downstream product choices feel off.
Afterall, a tool’s UI is just a representation of the database schema and so is the API. If the database model is designed for B2B then companies are a first-class citizen. This means that entitlements are built into features and companies, and that feedback is associated with features. That way everything in the product becomes simpler—the UI, the API, etc.
Then it feels like it’s a product that’s built for you, because it is. This means you don’t have to resort to workarounds, like awkwardly fumbling around with lists of UUIDs (as I’ve seen on way too many customer calls already). Instead, the defaults are intuitive and it just works. No yak shaving required.
So that’s why you should use feature flagging designed for B2B SaaS. But what are some concrete examples of how Bucket delivers on that?
Company as a first-class citizen
You obviously care deeply about your users, but in B2B SaaS you typically roll-out beta features to companies instead of users. There are exceptions, but you generally want every user in your company accounts to have access to the same features.
This saves you the hassle of always aggregating users into groups before setting feature access rules. By having a native company object, with its respective logo, it also means you’re toggling features on and off for companies, not lists of user IDs.
Furthermore, Bucket has a dedicated company page on which you can see, which features that organization has access to and for how long. You can even override feature access rules on the company page if you want to give a certain company access to a feature outside of their subscription permissions.
Gate features based on subscription plans
In B2B SaaS, customers are typically on a subscription plan with certain feature restrictions and usage limits, or on a bespoke Enterprise plan. With Bucket’s company segments, you can easily create dynamic segments that map to your subscription plans.
For example, you can create “Free”, “Pro”, “Business” and “Enterprise” plan segments and use those to gate features relative to your pricing table.
To gate a “Pro” feature, you simply select the “Pro” segment and higher meaning “Pro”, “Business” and “Enterprise”. Whenever a customer upgrades from Free to Pro, they automatically get access to the feature.
Bucket’s company focus also makes it very easy to report on feature access, like who has access to what features and for how long. You can even export the data as a CSV to use with other data sources.
Fix customer issues fast
When rolling out beta features, you flag them with a Bucket feature key. On Bucket, that same feature key can also be used to track feature adoption and to gather qualitative feedback from early adopters.
Having adoption metrics and feedback in the same tool enables small teams to move faster and with more autonomy. As mentioned above, the key is to address customer pains quickly in order to make customers happy and to spend less time on context switching.
Adoption metrics and feedback on Bucket enables developers to release a feature and instantly get feedback from actual users.
It's dead easy to add the in-app feedback button. Just add requestFeedback and trigger it on a button.
All feedback can be sent to Slack as well as to the Feedback tab on the feature.
Similarly, you can also use track to measure feature adoption. This gives you a list of early adopters who you can reach out to for some in-depth feedback once they've used the feature a few times. Longer-term, you can check out feature adoption metrics to see how well the feature is being used.
On the face of it, all feature management tools do the same thing: enable feature flagging. But especially in B2B SaaS, it’s important to pick the right tool for the job—one that’s built specifically for your type of customer and the unique problems you have in building products for them.