Data Analytics in Farm Accounting
Chapter Four
Syllabus topic 3, "IoT, Data Analytics with reference to Farm Accounting."
Pages 7 to 8 of 110
In one line
Data analytics is the examination of the farm's own recorded data to find what happened, why it happened, what will happen next, and what should be done about it.
The four kinds, each with a farm question
This is the structure to use in an answer, because it is complete and it is short.
| Kind | The question it answers | On a farm |
|---|---|---|
| Descriptive | What happened? | Cost per hectare of cotton last season was Rs. 48,200 against Rs. 44,600 the season before |
| Diagnostic | Why did it happen? | Because irrigation hours rose 40 per cent after the June rain failed, and labour rates rose 8 per cent |
| Predictive | What is likely to happen? | On this sowing date and this rainfall forecast, the yield will be 14 to 16 quintals per hectare |
| Prescriptive | What should be done? | Sow soyabean on the two upper fields and cotton only on the lower three |
Descriptive analytics is what ordinary accounting already produces. The other three are what analytics adds, and each one needs more data than the last.
What it needs before it can do anything
Analytics has no source of its own. Every one of the four levels reads the records of Module II. This is the link MU's topic line is asking for, and it is the reason this chapter sits where it does.
- Cost per hectare needs the cultivation register and the muster roll.
- Cost per litre of milk needs the stock book for feed and the milk record per animal.
- A trend in labour cost needs several years of attendance registers and muster rolls.
- A yield model needs recorded area, sowing date, input quantities and harvest weight.
A farm with no records has no analytics, whatever software it buys. That sentence is worth writing in an answer, because it is the whole relationship between Unit I and Unit III.
Where it is actually used
1. Cost analysis by crop, field and animal. The commonest and the most valuable.
2. Yield prediction. From sowing date, variety, soil readings, weather and past yields.
3. Input optimisation. Precision agriculture applies fertiliser and water by the square metre rather than by the field, from soil-map data. The accounting effect is a lower input cost per hectare and a consumption figure that varies within a field.
4. Price and market timing. Historic price series against the farm's own storage cost, to decide whether to sell at harvest or hold.
5. Livestock management. Milk per animal per day against feed consumed, to identify the animals that do not pay for their feed.
6. Machinery and labour scheduling. Hours available against the work the crop calendar demands.
7. Risk and insurance. Loss probability by field, from rainfall and yield history.
Data Analytics in Farm Accounting
8. Credit assessment. Lenders now score farm loans from yield, area and repayment data rather than from land title alone.
Two ratios worth naming
Analytics on a farm is mostly the disciplined use of physical ratios. Two carry most of the weight.
Cost per unit of output. Total cost of a crop divided by quintals harvested. It is comparable across seasons and across farms in a way that total cost is not.
Output per unit of input. Quintals per hectare, litres per kilogram of feed, quintals per labour day, quintals per thousand litres of water. These say where the farm is losing, and money ratios do not.
What analytics cannot do
- It cannot create data. Garbage in, garbage out is not a slogan here, it is the normal failure.
- It cannot make a decision. It narrows the choice; the farmer chooses.
- It cannot see what was never recorded, and on most farms that is unpaid family labour, home consumption and produce given as wages.
- It cannot beat a model that does not fit the farm. A yield model built on Punjab wheat does not transfer to Vidarbha cotton.
- Correlation is not cause. Two things that moved together for four seasons will part company in the fifth.
The link forward
Analytics needs a number for the standing crop and for the herd, and it needs the same number every year or the series means nothing. Ind AS 41 is what fixes that number. The next chapter begins the standard.
The rest of this subject
These notes are cut from the University's printed syllabus. Open the syllabus itself, or the past papers, for the same subject.