I’ve been managing the Fitbit Feature Suggestion board for over a year. When I started, as with many online idea submission sites, it had fallen into “idea overwhelm,” more clearly depicted in Bill Johnston’s model below under step 2, “overwhelming backlog:”

This happens all too easily in communities that are very busy or where idea boards are sometimes launched without a clear process in place to tackle them. Overwhelming backlog usually leads to a stagnated and unmoderated idea space where customers rightly grow frustrated, and may be eventually better shut down.

But hold on… there is a way back from this scary online place! And there are a lot of ways to improve how we listen to customers and advocate on their behalf. I hope these insights are useful to anyone running an ideas board or anyone representing customer or community feedback and ideas, which should hopefully be all community managers.

1. Get clear about what your Product team wants

I left innumerable meetings with product managers having shared what our customers wanted, thinking I was being a stellar community advocate, wondering why our PMs weren’t more interested.

I realized, however, that our ideas board proposition is an open text box, which is essentially saying “tell me absolutely anything you want.” And then, I took all of that information, and gave it to Product. That volume of information was overwhelming, and the content was often irrelevant and unspecific.

After that I began asking “what exactly do you want to know from customers?” and when they gave me clearer ideas, I could find that information through various narrower methods such as surveys, polls, private online discussions, or even enlisting our UX and marketing team to help me. Things moved more quickly after that.

2. Set your community up to succeed by asking the right questions

It didn’t take long to discover that ideas submissions weren’t as useful for our product team as they could be because customers usually jumped straight to their own solution for what they wanted fixed or improved, rather than focusing on the problem they were having.

And there are many ways to skin a rabbit.

My role is to represent what customers want, but ideally we also need to understand why they want it — yet we aren’t specifically capturing that insight. Focusing on their solution also makes for a messy ideas board. For example, if customers want to be notified when their Fitbit was out of range of their phone, or if some want an email or push notification if they haven’t worn their device for 24 hours — both of these are just different ways to solve “sometimes I forget to put my Fitbit back on after I showered.”

If you’re lucky enough to be starting an ideas board from scratch, or have the budget for a redesign, talk to your product team and find out what exactly they need to know to help them analyze and evaluate customer ideas. I bet most of their requirements can be designed into the UI where customers share their ideas.

After a co-production workshop with PMs, we’re adding another box which asks customers “What is the problem you’re hoping to solve?”

3. Get clear about what you’re asking of your community

Below are different “crowdstorm patterns” as per Crowdstorm, the future of Innovation, Ideas and Problem Solving by Abrahamson, Ryder, and Unterberg (which I’d highly recommend).

They each outline what we want customers to do:

1. Search pattern: members submit ideas to the organization. There is limited interaction between members submitting ideas and experts primarily from inside the organization decide on what are best.
2. Collaborate: Members submit but also provide feedback on each others’ ideas and evaluate them.
3. Integrate: Crowdstorming is integrated deeply into business operations and usually makes its way to prototyping or production.

At Fitbit we predominantly use the collaborate model where users vote and comment on each other’s ideas and discuss them, and I’d like to move towards an integration model in the future.

Getting clear on which model you’re using helps direct the user behavior you want and also sets expectations — asking customers to help co-create a product as LEGO ideas or local motors is very different to asking for ideas which go into a closed idea box, like at Starbucks.

4. Design for the caliber of idea you want

A general rule of thumb in design is to make it as easy as possible for people to do what you want them to do. However, my advice is don’t make it too easy… because the quality of submissions can be so low that you’re spending a disproportionate amount of time trying to understand what the idea is and how it might work.

I started thinking about submission systems which are much more rigorous – like a funding application, job application, or the idea boards setting out to solve extremely complex problems

like the NASA big idea challenge. This site has a clear document outlining criteria and allows participants to upload videos.

As a result, we’re currently redesigning our idea submission page so it appears more like users are writing to the CEO of Fitbit to propose their idea. We will also allow a space for them to upload files or images, to encourage the addition of mock ups or more supporting documentation.

5. Limit the ideas you’re fighting for

Most idea communities will have some sort of taxonomy or labelling system. I used to isolate all the ideas under a label and present 50 or so to the relevant team. I’ve had much better success narrowing my focus and picking the top 3 suggestions in an area and focusing in on the next sprint ahead.

6. Integrate strong ideas into your project management system

I used to cluster ideas for each team at Fitbit in a huge Google Doc – and honestly, still partly do — but those ideas are not going to get any visibility or attention until they’re in whichever system engineering is using. At Fitbit that system is Jira — so now, all of the ideas that have been agreed to go straight into a Jira ticket where I can chase down updates. I am known as a Jira stalker and will add supporting customer data to tickets wherever I can: votes, comment count, page views, etc. This also helps increase community visibility within the organization amongst staff I don’t normally interface with.

7. Don’t be afraid to be a squeaky wheel

Your job as someone who represents what customers want is to be a squeaky wheel. This means asking challenging questions at staff meetings, getting face time with VPs and other execs, and internally educating every team that they should be using you as a resource to understand what customers want. I’ve presented at dozens of departmental meetings to make sure everyone knows that we have a full-time staff member dedicated to customer ideas. Remember that with staff churn, you need to keep presenting and reinforcing this message, especially if you’re trying to drive culture change.

8. Make it easy for Product to view ideas

Most idea communities have a subscription functionality where you can automate sending certain ideas to certain people. A few community managers I’ve spoken to use this to make sure engineering see new ideas, for example, but this doesn’t always work at scale. We receive more than 20 new ideas each day, and over 400 comments, and thousands of pageviews on our idea community.

Engineers and designers don’t have time to peruse the community boards. I now send a quarterly infographic to a wide distribution to make sure people are reminded in their inbox that customers are busy asking us for new features or improvements, and I’ve found a colorful infographic can improve click-throughs!

9. Make exporting ideas as easy as possible

The best custom development we made to our Lithium community platform was building a custom export tool, where I can isolate and export ideas by label, status, date submitted, language, and vote count. This made reporting significantly easier (and standardized) and allowed me to track our progress. For example, I can see how many ideas with over 500 votes we are considering developing, or what percentage of all ideas ever get more than 3 votes (the answer is two thirds!). It also allowed me to export voters/commenters who were interested in a specific idea and share those with our UX or Field Testing teams, who are often recruiting for customers to interview or provide feedback, further incorporating real customers into our development processes — a big community win.

10. Reduce idea duplication

Idea duplication is toxic to idea boards. This problem will likely never be totally resolved (as there are so many word variations (e.g.: device, tracker, watch, clock, etc.), but we’ve made progress by forcing users to search before submitting an idea. We then present them with any close matches based on keywords to see if there is a suggestions that already exists for them to build on.

11. Feature Suggestions can be an SEO boon

We used to make customers register on community to be able to view or vote on suggestions. We now only require registration for idea submission and made our idea community indexable on search engines. This has, unsurprisingly, significantly boosted traffic. Customers can see specific ideas show up in search results and go straight to that page. This has also decreased customer support contacts where users call in or email us to tell us about an idea.

12. Not every customer idea is a good idea

We all know as community managers that what customers say and do is different. Just because they’ve posted an idea doesn’t mean they’ll use it. Your PMs should be pushing back a bit to verify how many users this suggested feature might apply to, how often it’ll be used, whether it impacts retention, etc. I’m not advocating for blind development of all customer suggestions.

13. Find out the real reason PMs aren’t implementing more ideas

I wrongly assumed that our Product team didn’t understand the value of community and set about presenting to everyone who would listen about why customers are important. Surprise, they already knew! The issue (as is often the case) was more around prioritizing customer input against dozens of other requests. Understanding what other constraints my development teams were working under helped me conclude that without accountability, nothing would change. So, now we have officially made customer ideas a field which our Product teams are measuring against in their internal score cards.

14. Build a panel of real community members to advise you

One of the best decisions I made was to immediately find the top 10 customers who were posting or replying to ideas, and create a community consultation panel. I ran all major changes by them and they were integral to the process of improving our ideas board. For example, I’d ask them “Would this template response declining a suggestion annoy you?” or “How can we manage expectations around long development times?” or “Is it annoying if we block comments?” As a matter or principle, we should always be involving real community members in our design and processes before rolling out big (or even small) changes.

15. Give context to vote count

Have you ever had an engineer say “But it’s only got 100 votes. That’s not enough customers”? Customers that engage with ideas are likely to be super-users (have you ever bothered to go on an idea board of a product that you love?! Most people haven’t) and therefore representative of a much larger demographic. Work with your data team to extrapolate some sort of voting scale or context. If market research asks me to give them an idea of how popular something is, I’ll tell them it’s in the top 10th percentile of ideas from all time, and that gives meaningful context.

16. Be as transparent as you can about how idea integration works

One of the first things I did at Fitbit was to create a FAQs page where we explained the process of choosing ideas from the community. Confidentiality and resulting transparency is likely always going to be a challenge, but something is better than nothing. We also put together this infographic about how feature suggestions works, to provide more context to their idea journey.

17. Be mindful about declined ideas

Customers are never going to be happy if their idea changes status to “not currently planned” or whatever your equivalent is. I worked with our customer consultation panel to brainstorm how to improve this experience, and we decided to block comments that are considered inflammatory. We did allow customers to continue voting on ideas that were publicly marked as “not planned.” This way, I can continue to track demand and challenge decisions internally around why we’re not developing something if customers continue to show interest.

18. Develop a public archiving policy

Idea overwhelm is real for everyone — including customers! We archive ideas that have garnered less than 5 votes in 6 months in our English, German, French, and Italian communities. We have a lower bar for our Spanish and Japanese idea boards, which have lower traffic. If less than 5 customers vote on an idea, it’s unlikely to be a popular feature or worth the investment. I still store all these archived ideas on a private board so I can still run searches and exports and include them in data sets internally, but they won’t clog up the public boards.

In Conclusion…

These are just a few of the things I’ve learned whilst managing Fitbit’s feature suggestion board. I hope they’re useful to you, whether you manage a formal ideas board or if it’s part of your role is to represent customer feedback. I often ask myself, ‘have I ever been so passionate about a product that I’ve bothered to find their idea board and submit a suggestion to that company?’ – and honestly, the answer is no. So, I deeply appreciate these customers and their taking the time and energy to share their ideas with us to help us create the best products possible.

As Richard Millington says in The Indispensable Community, “Ideas are interesting, but the insights are indispensable. Insights can validate a company’s way of thinking, identify new opportunities, or change how they think about a problem. Even bad ideas can yield good insights.”