Which comes first on a conversion feasibility screen, the test fit or the hard cost range?

I didn't know a test fit was a thing you could buy separately. Reading the rest of this thread to figure out which one I'd even be able to ask for.

Lantern, the systems walk comes first, with one condition. The GC has to be told what the target looks like in rough terms, so studio heavy versus one bed heavy, and roughly where the corridor wants to run. Give them that and the range tightens up a lot. Send them in cold and you get 200 to 400 a foot and a shrug. What matters on the first walk is the stuff no drawing changes. Slab thickness and whether it is post-tension, because if it is, coring happens where the tendons allow and the plumbing routing gets ugly fast. Existing service size at the main. Whether the shafts can carry new risers, or whether new ones have to be built through floor area that was going to be rented. Floor to floor height, because once a ceiling drops for horizontal runs on a 9 foot 6 slab to slab, the units sit at 7 foot something and they feel like a basement. None of those answers move based on the unit layout. All of them can kill the deal. A test fit tells you what you would like to build. The walk tells you what the building will let you build, and that constraint is worth learning before anyone pays to draw around it.

Across a wide screening pipeline, most candidate buildings never earn either spend. Both of these belong at stage three. Stage one is the photos and the footprint. Depth from window line to core matters, because past about 30 to 35 feet on a double-loaded office plate you are generating interior space nobody will rent as a bedroom. Window spacing against a plausible unit width comes next. Whether the building is a single story box with no second floor to sell changes the whole math. Stage two is the county records and the zoning text. Residential permitted by right or not, parking ratio, and anything recorded against the parcel. Stage three is where lantern's question starts, and by then nine out of ten are already dead. So the honest position is that arguing test fit against cost estimate is arguing about the tenth building. Girder's list is still the better use of a first site visit, since layout guessing can be done off a floor plan and a post-tension slab cannot be guessed at.

The order depends on which variable in your model has the wider range on this specific building.

If your unit count could plausibly swing 30 percent, the test fit goes first, because unit count sits in the numerator of your revenue and in the denominator of your per-unit cost. Both move at once. That's common on deep-plate office where the answer to how many units you get is genuinely unknown until somebody draws it. If your unit count is close to determined, say a shallow plate with regular punched windows where anyone competent lands within a unit or two of the same answer, then the layout is not where your uncertainty lives, and paying to resolve it buys you little.

On most conversions the wider range is hard cost, and the widest single item inside hard cost is plumbing and structural work to accommodate it. So the sequence a lot of capable operators run is a short engineering opinion first, aimed narrowly at slab, risers, shaft capacity, and incoming service. That's a scoped consultant hour, not a full report. It resolves feasibility as a yes or no. Then the test fit, once you know what routing the building tolerates, because the layout should be drawn against real riser locations rather than redrawn later when they turn out to be somewhere else.

One caution on the cost side. A GC range without a layout is a range on a building type, not on your building. Treat it as a screen that can kill a deal and not as a number that can approve one.

What you're actually selling, lantern, is a decision rule that tells the client which of those two uncertainties is larger on the building in front of them. Sell that ordering logic and the deliverables follow.

I'm still working out what my first deal costs, so this thread is useful in a way I didn't expect. Both of these are real line items before I own anything, and they don't come back if I walk.

What I want to know from the people who've done it, and lantern maybe you've seen this across clients, is what the total pre-contract screening spend looks like on a conversion that gets killed. If the systems walk is 1,500 and a scoped engineering opinion is another 3,000 and a test fit is 2,500, I'm at seven grand on a building I never buy, and if I have to do that four times to find one that works, that's a real number I need in my budget line before I start.

Rena-expert's ordering makes sense to me exactly because it front-loads the cheapest kill. Spend the small money on the question most likely to produce a no.

Coming at this from the operating side rather than the conversion side, the thing both camps tend to skip is unit mix versus what the submarket actually absorbs. A test fit will happily produce twenty two units at 480 square feet with one window each, and that yields well on paper while leasing like a rock in a market that wants two bedrooms. The rent survey and absorption read should come before either the test fit or the hard cost estimate, not because the engineering does not matter, but because the engineering question is what will this cost and the market question is what will anyone pay for the result. A clever floor plan with the wrong bedroom count is still a vacancy problem two years later. Between systems and drawings, systems tends to be the safer first dollar, since the drawing cannot argue with what the building actually needs.

Scope creep on conversions almost always comes in through the ceiling, which is an argument for the engineering walk. Every horizontal run you didn't plan for eats headroom, and every inch of headroom you lose gets bought back with structural work or a smaller unit.

But I'll push back on girder a little. The systems walk gets you a no. It rarely gets you a defensible yes, because the pricing on a conversion moves enormously depending on how many wet walls you end up with, and wet wall count is a layout output. I've had a GC number come in at 240 a foot against a rough concept and land at 305 once the actual plan stacked bathrooms badly and we were coring in eleven more places than the concept implied. That delta wasn't inflation or surprises in the field. It was the drawing changing the scope after the price was given.

So the sequence that actually works for me is walk, kill or continue, then test fit, then re-price against the test fit and treat the first number as thrown away. Two cost exercises. It feels wasteful and it's cheaper than being wrong.

What gets a conversion looked at seriously by credit is a cost number tied to a specific drawn plan and a third party who signed it. A GC range against a concept doesn't survive the first question, because nobody underwriting can tell whether the range covers the actual building or just a category of building. So whatever order runs internally, both deliverables end up existing before anyone lends real money against the project. The screening question is just which one gets bought while it's still possible to walk for free. Draw structures, holdbacks, and how a lender treats contingency on a conversion vary a lot by lender and by program, so those terms are worth confirming in writing with whoever is actually underwriting rather than assuming the standard construction treatment applies. The two-price approach is the version that shows up most often on deals that get funded. The first number decides whether to keep spending. The second number is the one anyone underwrites.

The service angle here is interesting to me because lantern is describing a product with a pricing problem, not just a sequencing problem. If the first deliverable is the one most likely to kill the deal, the client experiences your best work as the thing that made them lose money. That's a hard sell to repeat customers unless you frame the fee as insurance against the bigger spend.

The operators in this thread who've done it seem to want a scoped engineering opinion, a kill or continue decision, and a test fit afterward. That's three separate small engagements rather than one report, which is probably a better business anyway since the client can stop at any point and still feel like they got the thing they paid for.

Lantern, do your clients want the recommendation or the raw findings? I'd guess the ones who've done conversions before only want the findings and the ones who haven't want you to tell them yes or no.

Reading this thread as someone newer to the vocabulary, what stands out is that a post-tension slab is apparently a deal killer and riser capacity decides everything, which is not where the intuition would naturally go. One thing worth flagging. If the systems walk can kill a deal but can't approve one, and the test fit can't be priced properly until the systems are known, then the first spend is always the one that produces bad news and never good news. That may just be how conversions work, paying repeatedly for permission to keep paying, until one building runs out of reasons to say no. The engineering walk holds up on that strength.