Seriously though, @staff and @changes -- You have a literal goldmine of user insight right here, in your userbase.
Why should you even read this? Because I've literally just accepted a gig to help the execs of a major lending institution rework its tech-centric product management. Because the culture of their company has "forgotten how to think about their customer needs" and are getting their lunch eaten by competitors who DO. And they've asked me to help.
So:
Open up a form for "how should we make tumblr profitable?"
Fields: Short Description, Long Description, Funding Source (Who will pay?), Business Justification (Why will they pay?), Example (Where else do you see examples of this model working?).
Short Description is the only required field. Messaging on the form that "clearer and more complete ideas have a greater chance of review."
Basic genAI semantic checking on the form input to validate that whatever was entered into the form is at least comprehensible and relevant to the question, not keysmash or lorem ipsum.
Partially-automated pipeline for submissions. Workflow states = Received, Slush Pile, For Review, Reviewed, Plausible, Proposed, Feedback Received, Pilot, Rollout, Done.
Automated triage: All 5 fields filled -> advance to "For Review" state. Short Description is only field filled -> drop from the pipeline. Aggregate together with similar submissions, generate a keyword histogram. More than 1 but less than all 5 fields filled -> "Slush Pile."
Start working the "For Review" submissions. If it's a business revenue idea with justification WHETHER THE IDEA IS PLAUSIBLE OR UNLIKELY, it goes to "Reviewed" state. If it hasn't been reviewed, it stays in "For Review." If it's a plea, wish, or plaint, it also drops out of the pipeline for key phrase analysis/sentiment analysis in the aggregate along with similar entries.
Start working the "Reviewed" submissions. Is it plausible? Could you test it in a lightweight way to determine its likely ROI if rolled out? -> Advance it to "Plausible." If not, compare it to your (by now, rather large) set of keyword/sentiment analysis and see if it matches any emerging themes.
Bring the Product people in to start working the "Plausible" ideas. THIS is where you are really seeding the future revenue capability of the site with priceless intel gleaned from your userbase. Can some ideas be combined? Do some groups of ideas suggest a single service with multiple tiers? Might some work, but for a different market/payer than originally envisioned? Are some good bets as-is? Sketch out some lightweight roadmaps for the ones that make the cut. Write them up as brief proposed changes. Create polls. PUT THESE IDEAS IN FRONT OF users, advertisers, network partners -- whoever has been identified as the payers for the idea. These ideas move to -> "Proposed."
AT THE SAME TIME, also release your analysis of all the aggregate submissions that didn't make it this far. What are the themes? What did you -- and Product -- find surprising? Is there anything way off about very popular themes; ideas that seem to have a lot of support but are completely divorced from reality? Say so. Explain why. Start the dialogue. Your userbase learns more abt what it takes to stay in business. You keep an open channel to your biggest asset: your userbase.
Collect the feedback on the "Proposed" ideas. These move to -> "Feedback Received."
Some will die here, the polling showing there's no market for the feature. Some will need to be reworked based on feedback. But you'll start accumulating some good "bets" for future revenue generators. Get Product, Engineering, and MarComm in a huddle: Estimate some time/cost to implement, time until ROI, market penetration, and growth curves. Pick the best.
Time to test your best bets. All the brainstorming, all the polling, it don’t mean jack compared to real numbers. Do LIGHT LIFT Pilots. They don't have to be optimized. They don't have to be for all users. They don't even have to be bug-free! (But users should be able to report errors.) Make it very clear that this is a PILOT. Run these features at a small fraction of the proposed final cost. Because you need to charge real money or it's not a test, but you can't charge full price, because it's just a pilot.
Compare the results against your estimates from the previous step. For the Pilots that show promise and meet or exceed their projections, -> Rollout. For those that don’t, communicate the results back to the userbase. They WANT to help you.
Roll out the new feature/service/capability. Across the board, as robust and bug free as you're able.
Repeat.
Note this is CONTINUOUS FLOW, not batch-and-queue. You're continually triaging, evaluating, communicating, projecting, piloting, rolling out...... all while incorporating incredibly valuable insights from your userbase, AND educating your userbase at the same time.
Really, why be indifferent to -- and thus be perceived as antagonistic to -- your users, when you could PARTNER with them and GROW?




















