In today’s competitive software market, companies are constantly racing to develop innovative features that engage, delight, and retain their users. All kinds of organizational resources, from engineering manpower to marketing budget, get allocated to feature development and release.
However, this rush to build new things and get them out the door often causes older features to fall by the wayside. It becomes very easy to sacrifice long-term, strategic gains for short-term profit. Plus, the breakneck pace of new feature releases means teams don’t have time to evaluate how existing features are being used (if they’re being used at all).
Pendo, the company behind ProductCraft, isn’t immune to these challenges. As Pendo’s data scientist, my focus is on optimizing user delight while driving the company toward its growth goals. To do that, I need to understand how users are interacting with our features, both old and new.
I recently made it a top priority to dig into the data on feature utilization, which inspired our team to build a tool to measure feature adoption. This tool helped us answer several key questions:
- Which of our features have the highest (and lowest) average daily adoption rates?
- How do different features compare in terms of adoption rates?
- What percentage of features are rarely or never used?
Our Data-Driven Findings
Once we put our feature adoption tool to work, we made some interesting discoveries.
Across customer products, we found that on average, 80% of click volume came from as little as 12% of features. This is much lower than the 80/20 Pareto rule commonly hypothesized for software usage. Additionally, as much as 80% of features in the average software product were rarely or never utilized.
By our estimate, publicly-traded cloud software companies invested up to $29.5 billion in the development of these features alone (please refer to this paper for more on our research methodology and findings).
For your company to achieve positive ROI, the entire team should emphasize the evaluation of your current feature utilization rates, then apply that data to your product planning and development roadmap.
Here are five things my team found to be critical when assessing and improving our own product’s feature adoption rates.
Measurement and Benchmarking
Regularly establish which are your most- and least- adopted features, and see if these lists match your expectations. Are there sharp changes in the feature leaderboard by week/month/quarter? Do any of the entrants in the lists surprise you or other stakeholders in your company?
Validating your assumptions with data is important not only as a gut-check but also for accurate benchmarking. With solid initial benchmarks, you can continue to build a sustainable strategy.
Go through your list of product features at a regular cadence (at least yearly). If you have features that aren’t being used at all, dig deeper to find out why. Are they working as intended, i.e., are there any technical errors you may need to resolve?
Whether or not you should keep those least-adopted features will likely hinge on the trade-off between the costs of maintaining them and the revenue generated from their use. If the balance tips in favor of the former, it’s time to prioritize other product features.
If you’re about to release a new feature, set a clear, achievable adoption rate target so you can measure its initial success. You might also identify an existing feature that has the potential to generate greater customer value if its adoption rate improves. Select an adoption-rate growth goal that you can track frequently and be accountable to.
Do you want to increase the adoption rate of a specific feature? If so, dive deeper into the feature’s usage behavior. Who uses it most? How does the usage behavior of this audience compare to that of a frequently-used feature? Is it hidden in your product, and if so, could you programmatically target an in-app guide to promote the feature to the desired user group?
Perhaps you could survey your users for feedback on this feature and other potential areas for improvement. NPS text sentiment data could also be a valuable resource for discovering why users are experiencing friction when using a given feature.
You may have more than one strategy available for enhancing the adoption of a given feature. In that case, run a few experiments to determine the best option.
Perhaps an enhancement of the feature (re-visualization, the addition of an up- or down-trending metric, etc.) would drive up the adoption rate. Maybe a UX re-organization of your product would enhance feature visibility. Analyze the paths taken to the target feature by its successful users. Can you increase user engagement at key moments in those paths?
Test these strategies with smaller audiences when possible. In certain contexts, you may even use a simulation study to help you predict the outcomes of these experiments.
Before running off to ship the latest and greatest product update, take a moment to evaluate your existing features. The data you dig up may surprise you! With some investment, your product’s “hidden gems” will continue to drive user engagement and satisfaction.