Even with a random sample, 22 weeks on average would mean more than 22 weeks for some people and less than 22 weeks for other people. So even if you don’t do anything (or just the bare minimum, to show that you did something), the latter would pay you.
It’s like making a deal with someone that you will help them predict the winning lottery numbers. If they win, they will pay you X% of the money; if they don’t, the advice was free.
(I don’t think this is exactly what the author is doing, I suppose there is actual help provided, but the calculation of the provided value seems misleading.)
Oh no, I assume you do help! At least I have no reason to suspect you.
My objection is that the pricing strategy “you only pay if it takes you less than average time” sounds like a proof of something, when actually it’s not. That is, hypothetically, if you did nothing—which I don’t think is your actual plan—and only took money if the time is below the average… you would still be making money. Of course, if your help works, you make more money.
I am not accusing you of anything, but there are people out there who make money the way I described: promise to improve a probabilistic outcome, do nothing, collect money only in the lucky case. (One case I read about in newspapers, another one I heard rumors of.)
My point is: returning money if the result is not above average is not a guarantee of successful intervention. (Shminux said it’s because your sample is probably better than average. I said it works even for an average sample.) You asked to “critique this pricing model”, this is the kind of nitpicking you got. From the perspective of average customer, I suppose your pricing model sounds great.
Even with a random sample, 22 weeks on average would mean more than 22 weeks for some people and less than 22 weeks for other people. So even if you don’t do anything (or just the bare minimum, to show that you did something), the latter would pay you.
It’s like making a deal with someone that you will help them predict the winning lottery numbers. If they win, they will pay you X% of the money; if they don’t, the advice was free.
(I don’t think this is exactly what the author is doing, I suppose there is actual help provided, but the calculation of the provided value seems misleading.)
I’d a love a little more detail here. What makes you think I’m not providing help?
Oh no, I assume you do help! At least I have no reason to suspect you.
My objection is that the pricing strategy “you only pay if it takes you less than average time” sounds like a proof of something, when actually it’s not. That is, hypothetically, if you did nothing—which I don’t think is your actual plan—and only took money if the time is below the average… you would still be making money. Of course, if your help works, you make more money.
I am not accusing you of anything, but there are people out there who make money the way I described: promise to improve a probabilistic outcome, do nothing, collect money only in the lucky case. (One case I read about in newspapers, another one I heard rumors of.)
My point is: returning money if the result is not above average is not a guarantee of successful intervention. (Shminux said it’s because your sample is probably better than average. I said it works even for an average sample.) You asked to “critique this pricing model”, this is the kind of nitpicking you got. From the perspective of average customer, I suppose your pricing model sounds great.
Got it—makes sense. I set up contingencies where the client can terminate the contract (and vice versa) to solve against this.
Not a perfect solution though.
Really appreciate the thoughts here! It’s helpful.