Hi there! I always like to reiterate -- there is never a RIGHT answer, mostly in consulting interviews, the interviewer is trying to understand HOW you think and if the LOGIC you apply can be rationalized or reconsidered.
The approach is okay, but can be strengthened on what "point of view" you are taking to answer this question. It also depends on what data is available to you. You may have to readjust based on what data is available, because some of these statistics would be hard for you to estimate yourself (for instance, if the interviewer does not provide the number of pumps average for a station).
Suggestions to improve your approach:
- Point of view: the (avg. spend x frequency) perspective of your estimation is based around a customer (you are essentially estimating number of trips) but then this is mixed with number of pumps, which is not customer-based. Try to be linear with the thinking on the "units" that you are estimating
- If you took a customer based POV, one option is to do avg customer spend per year (easy for you to estimate a cost based on your own experience) x % of population with a car or driving (also an easy estimate) x total population size / number of petrol stations (can estimate with a ratio to population)
- Alternatively if you mean frequency = transaction count per day per pump, then it makes sense because everything is then pump based. Change the language to be more "pump based" POV (avg. spend --> average transaction value at pump)
- Remember that operational efficiency is never 100%. Inherently buried within your formula for petrol revenue is the assumption that all pumps will be working at all times. You may want to take a reasonable discount on broken equipment or repairs time (whatever the %, just make sure you can rationalize it eg "When I get gas, it seems that 9/10 times the machine is working, but 1/10 trips I make, I have to switch pumps because I pulled up to a broken one")
Some triangulation methods/follow ups:
- You are sizing bottom-up through this method (number of pumps, .... using smaller data points sizing up to the large aggregate annual revenue). Always, the opposite side of the coin to triangulate, is top-down (taking aggregate, higher level numbers to boil down to the answer)
- For example, [ total amount of petrol litres sold in UK per year x avg cost of petrol litre / number of petrol stations in UK ]
- Interviewer could ask for a more thorough answer on the convenience store revenue
- Interview could ask for an estimate on how much the annual revenue could vary (eg. sensitivity on the number), what is the biggest risk/underlying assumption you have made, etc
Hope this helps! Please feel free to reach out if any questions!