Deciding what to build next is the hardest part of the job. Harvestr gives you a flexible scoring system so that decision rests on evidence rather than the loudest voice in the room.
The Product interface
The Product section gives you a central view of every Discovery you're working on. It serves two purposes: reviewing your backlog, and deciding what to build next.
Your prioritization criteria
Two criteria exist by default on every Discovery:
Feedback volume — how much feedback is currently linked to it
User and company segments — the segments of the customers behind that feedback
Everything else is yours to define with custom fields. The types that matter most for prioritization:
Field type | Good for |
Numeric | Users impacted, days of work, expected revenue |
Rating | Impact, effort, confidence — anything on a 0 to 5 scale |
Cumulated ARR, count of Enterprise customers, feedback flagged as churn risk — computed automatically | |
Score | Combining the above into a single comparable number |
Create them from Manage fields in the top right of the Product section.
Build a prioritization score
Once your fields are in place, a Score field combines them into one number you can rank by. Harvestr offers two models: RICE and weighted average.
Create one the same way as any other field, selecting the Score type.
RICE score
RICE stands for Reach, Impact, Confidence and Effort. It weighs what an opportunity could deliver against what it costs to build.
RICE = (Reach × Impact × Confidence) ÷ Effort
In Harvestr, you map each element of the formula onto one of your Discovery fields.
Save the mapping and the score is calculated across all your Discoveries.
Effort sits in the denominator, so higher effort lowers the score. Make sure the field you map there measures cost, not ease — mapping an "Ease" field would rank your hardest work highest.
Weighted average score
The weighted average is the flexible option. It lets you:
take as many criteria as you want into account
give each one a weight reflecting how much it matters to you
end up with a final score that's always a number between 0 and 100
Weights
Assign each field a weight from 0 to 100, based on how much it should count.
The Score influence column shows how much each field weighs relative to the others. A field with 80% influence accounts for 80% of the final score.
Inverted criteria
Toggle Inverted on a criterion and it counts against the score: the higher the value, the lower the result. This is how you handle effort, cost, or risk in a weighted average.
How the calculation works
Harvestr always gives you a number between 0 and 100, so scores stay comparable.
To get there, each field value is first normalized: divided by the highest value that field holds across all your Discoveries.
Say you have a field for expected revenue. One Discovery is worth $10,000, and your highest anywhere is $100,000. That Discovery scores 10% on this field.
Each normalized value is then multiplied by its weight, and the results summed — giving a final score between 0 and 100.
Normalization is relative to your current backlog. Adding one Discovery with an unusually high value rescales everything below it. That's intended — scores answer "how does this compare to everything else on my plate" — but it's worth knowing before you wonder why a score moved on its own.
A worked example
Three criteria A, B and C, weighted 20%, 30% and 50%.
A | B | C | |
Discovery #1 | $10,000 | 10 | 1 |
Discovery #2 | $5,000 | 15 | 3 |
Discovery #3 | $3,000 | 20 | 2 |
Discovery #4 | $10,000 | 50 | 4 |
Discovery #5 | $15,000 | 70 | 1 |
The score for Discovery #1:
[20% × $10,000 / MAX(A) + 30% × 10 / MAX(B) + 50% × 1 / MAX(C)] × 100
= [20% × $10,000 / $15,000 + 30% × 10 / 70 + 50% × 1 / 4] × 100
= 30
With Inverted switched on for criterion A:
[20% × ((MAX(A) − $10,000) / MAX(A)) + 30% × 10 / MAX(B) + 50% × 1 / MAX(C)] × 100
= [20% × (($15,000 − $10,000) / $15,000) + 30% × 10 / 70 + 50% × 1 / 4] × 100
= 24 (rounded)
Putting it to work
With your fields and score in place, filter and sort to extract the subset you care about. Save the combinations you use often as Discovery views.
Three prioritization questions and how to answer them:
Trending requests from large accounts
filter on your "Feature request" tag
filter on the "Large accounts" segment
filter on last feedback within the past month, so you only see topics that are still live
sort by feedback volume
Quick wins
rate each Discovery on impact and ease, 1 to 5
build a weighted average giving each 50%
sort by that score — high impact and high ease rise to the top
Revenue-weighted priorities for a segment
filter on the segment you're focused on, "Power users" for example
build a score combining feedback volume, impact, effort and a Rollup field on expected revenue
sort by score
Sorting by feedback volume and sorting by cumulated revenue rarely give the same answer. The gap between those two lists is usually the most useful conversation your team will have this quarter.
