Most deals don't. That's normal, and the point of running the numbers early is to find that out in ten minutes rather than three weeks and a survey fee.
Here's the order I'd do it in.
Start with what it sells for, not what it costs
Everything hangs off the finished value. Look at what similar properties have actually sold for nearby, using Land Registry rather than asking prices. Get this wrong by 10% and it usually takes half your profit with it, so it's worth more time than anything else on the list.
Then the build
Cost per square metre will get you close enough at this stage. If the deal only works at the very bottom of a sensible range, it doesn't work. You're not going to be the one person who gets a full refurb done for less than anyone else.
Then the costs people forget
Stamp duty at the additional property rate, which is 5% on top of every band. Professional fees, around 12% of the build. Contingency, at least 10%. Finance. Selling costs. On a longer job, council tax, insurance and utilities while it's sat there.
Then check the margin
Profit divided by the finished value. Around 20% and you've got room for things to go wrong. Between 12% and 20% it can work but you're relying on a clean run. Under 12% there's no buffer at all.
Then work out what would break it
This is the bit people skip. Take 10% off the sale price and see what's left. Add 15% to the build and see what's left. If either one on its own kills the deal, you're not buying an investment, you're buying a bet on nothing going wrong.
Then the things that have nothing to do with the numbers
Flood zone. Conservation area. Article 4 if there's any chance of an HMO. Whether similar schemes nearby have been getting planning or getting refused. A deal that stacks beautifully and can't get consent isn't a deal.
The bit that isn't a calculation
Time. A 20% margin over eight months is a good job. The same margin over two years, with your cash tied up the whole time, is a lot less exciting. Nothing in the maths tells you this, so you have to think about it separately.
And if you find yourself lowering your target margin to make a deal work, that's the deal telling you something. Everyone does it at some point. It's almost always the one you regret.
Avalor runs all of this from a postcode and a purchase price, including the planning and flood checks, and shows you how far the sale price or build cost can move before the deal stops working.