Aquaculture app ROI: what it is worth per pond, per crop
A farm management app does not grow shrimp. It shortens the distance between something going wrong in a pond and somebody knowing about it, and that distance is where the money is. The question worth answering before any rollout is what that shortened distance is actually worth per pond, per crop, in numbers a farm operator can check.
What follows is a model, not a result. Every assumption is stated so it can be argued with, and the figures below are modelled outcomes rather than measured performance from any particular farm.
The baseline pond
The model uses a one-hectare pond running a 120-day crop, producing 6,000 kg at 75 percent survival with a feed conversion ratio of 1.45. Shrimp is priced at 4 dollars per kg and feed at 1.20 dollars per kg.
That baseline is deliberately unexciting. The FAO cultured species profile for whiteleg shrimp puts feed conversion in intensive systems at 1.4 to 1.8, so a 1.45 starting point is a normally run pond rather than a struggling one. The model has to earn its return against a competent operation, not a broken one.
What the app can plausibly move
Two variables, both of them consequences of noticing things sooner.
Survival. Stocking, water quality, and disease pressure decide how much of what went in comes out. An app cannot change any of those. It can put a dissolved oxygen reading in front of someone the day it drifts rather than at the next visit.
Feed conversion. Feed is the largest controllable cost in the cycle. Consistent daily feeding records plus sampling data are what make it possible to feed to actual biomass rather than to a schedule.
The model, per one-hectare pond
| Outcome | Conservative | Expected | Strong |
|---|---|---|---|
| Survival | 76.5% | 78% | 81% |
| Feed conversion ratio | 1.42 | 1.38 | 1.30 |
| Harvest | 6,120 kg | 6,240 kg | 6,480 kg |
| Gross benefit per crop | 492 dollars | 1,067 dollars | 2,251 dollars |
The expected case breaks down as follows. Survival moving from 75 to 78 percent lifts the harvest by 240 kg, worth 960 dollars. Feed conversion improving from 1.45 to 1.38 means total feed falls from 8,700 kg to 8,611 kg despite the larger harvest, saving 89 kg of feed worth about 107 dollars. Together, 1,067 dollars per crop.
Note the shape of that result: roughly 90 percent of the benefit is additional shrimp and only 10 percent is feed saved. Feed conversion is the number farm managers talk about, but on these assumptions survival is what actually pays. A three point survival gain is worth nine times the feed saving.
Why the strong case should not be the business case
It is tempting to model against published trial results, where controlled ponds report considerably better survival and feed conversion than any of the columns above. That comparison is misleading. Those trials typically involve controlled stocking, monitored water chemistry, and often sensors or automatic feeders. An app with none of that hardware attached is a record-keeping and alerting layer over human judgement, and it should be expected to capture a fraction of a controlled trial's improvement rather than all of it.
The same caution applies to sample harvest figures from a well-run demonstration pond. A pond that returned 90 percent survival at 1.10 feed conversion is a useful illustration of what good looks like. It is not a rollout baseline, and modelling a fleet against it will produce a business case that misses.
Scaling to 100 ponds
At two crops a year and 70 percent active adoption, which is a realistic rather than optimistic assumption for field software:
- About 33.6 tonnes of additional shrimp per year
- Roughly 12.4 tonnes less feed consumed
- Around 149,000 dollars of annual gross production benefit
- Approximately 470 technician hours returned, assuming eight visits per crop and 25 minutes less reporting per visit
At full adoption the production benefit rises to roughly 213,000 dollars a year. The gap between those two numbers, about 64,000 dollars, is the price of the 30 percent who stopped logging. That gap is the entire adoption problem expressed in currency, and it is usually a bigger lever than any feature on the roadmap.
The conditions the whole model rests on
Every figure above assumes behaviour that does not happen automatically. The targets worth writing into a pilot, and measuring weekly:
- 70 percent active usage
- 75 percent daily feeding-log completion
- 80 percent same-day water-quality logging
- Technician response to red alerts within 24 hours
Miss those and the model does not degrade gracefully, it collapses to a different category of return. Without consistent logging there is no biomass estimate, so feeding stays on a schedule and feed conversion does not move. Without same-day water-quality entry the alerts arrive too late to act on, so survival does not move either. What survives is the administrative saving, the technician hours, which is real but is roughly one tenth of the case.
This is why adoption metrics belong in the pilot design rather than in a post-rollout review. Downloads and logins predict nothing. Completed feeding logs predict the entire return.
What this means for building one
The design consequences follow directly from the model. If the return depends on daily logging by people standing at a pond, then install size, offline capture, and the number of taps to record a feeding are not user experience preferences, they are the financial model. An app that requires a live connection at a site with no coverage does not produce a smaller benefit, it produces the administrative tenth.
We build aquaculture software on that basis: offline-first capture, per-species safe ranges attached to water quality readings, and feed conversion calculated from what was actually logged. The offline-first pond management app we delivered for a feed and farm-services group replaced paper culture record forms across several country subsidiaries in six languages, because the alternative was a system the field abandoned in week one.
If you are building a business case for one, take the model above, substitute your own shrimp price, feed cost, and baseline harvest, and be honest about which adoption column you will actually hit.
