Attached garage conversion versus detached cottage on the same lot, comparing an ADU decision beyond the per-dollar math

Studying for the license exam right now and one thing that stuck: in a neighborhood where every house has a garage, taking yours away shows up in the adjustment grid. How much depends on your market. Ask an appraiser directly, some will do a consult for a few hundred dollars.

Nine to eleven weeks on a garage conversion usually means fourteen to twenty in practice. These rarely come in on the quoted timeline once the slab is opened and the grade turns out wrong for a habitable floor. Attached garages usually sit a step below the house floor, and getting drainage and moisture right under a new slab is where the change orders live. Ask what happens if the slab has to be replaced instead of topped.

The third unknown is not an unknown, it is a document nobody has finished reading, and it is the one of the three that can void the project. Start with the ordinance's definitions section. Most ADU ordinances define separate categories, something like internal, attached, and detached, and apply different rules to each. Size caps, setbacks, parking, and owner occupancy are frequently written per category. A converted attached garage may fall under internal or attached depending on whether it sits inside the building envelope, and that classification alone can change the maximum size and whether owner occupancy is required. If owner occupancy language shows up in a section about something else, check whether it is a general condition of the permit rather than a rule tied to one category. How the ordinance applies to a specific lot is a land use question for an attorney where the language is genuinely ambiguous, but a zoning administrator will often give a written interpretation for free when cited to the section numbers. Then check three documents the ordinance will never mention: the deed for restrictive covenants, any HOA declaration, and the title policy exceptions. Private covenants can prohibit a second dwelling even where zoning allows it, and the city will not flag that when it issues the permit. Plenty of mid century subdivision covenants still carry a one dwelling per lot clause. It is also worth pulling the homeowners policy and reading how it treats a rental unit on the premises, in writing, from the carrier, before committing to either budget.

Warrant, the covenant point is the thing I'd never have thought to check. Anchor, is your neighborhood a platted subdivision or older infill? Makes a difference to whether there's a declaration at all.

The obvious question to ask first: does either number include the utility work? Separate meter, sewer connection fee, tap fees? Those are city fees and they can be five figures on their own.

Vellum asked the right thing. Tap and impact fees are the line item that shows up after the contract is signed, because the contractor's quote is for construction and the fees are separate. In some jurisdictions a detached unit triggers a full sewer connection fee and an internal conversion doesn't, since it's on the existing lateral. That could be $8,000 to $20,000 of swing between the two options and it isn't in either number as posted. On construction: the garage conversion at $96,000 for 440 feet is $218 a foot, and the cottage at $178,000 for 700 feet is $254 a foot. Both are credible. Conversions look cheap per foot until you price the floor, the insulation in an uninsulated shell, and a fire separation wall between the unit and the house, which many codes require on an attached unit. Ask specifically whether the quote includes rated assemblies.

Scope creep on garage conversions is a genre. My worst one: opened the wall to run plumbing and found the header over the garage door was carrying more than it should have been, and we were into a structural fix that nobody bid. Anchor, whatever you pick, hold 15% back in your own budget that the contractor never sees.

Parking is what decides it for most owners weighing this choice, and that pattern holds across neighborhoods where street parking is already tight. If a street has no overnight parking and the garage disappears, the result is a tenant with a car and nowhere to put it, and that shows up as complaints and turnover, not as a line in a spreadsheet.

Who's managing it? A detached unit with its own entrance and its own meters is close to a normal rental. An attached one bedroom sharing a wall with your bedroom means you're the maintenance person and the noise complaint department at the same time.

Read the insurance angle warrant raised more carefully than the rest of this thread. Most standard homeowners forms have limits on rental activity on the premises, and some carriers will move the policy to a different form entirely once there is a separate dwelling unit. Get the premium change quoted before you choose, in writing. A $1,400 annual increase is enough to turn a marginal spread into nothing.

Anchor, your per-dollar comparison is comparing gross rent to construction cost, which isn't a return. Option A: $1,325 times 12 is $15,900 on $96,000, or 16.6% gross. Option B: $21,900 on $178,000, or 12.3%. Now take out the fees girder mentioned, fathom's insurance, the property tax reassessment your county will do on new construction, and 8% for vacancy and turns. The gap narrows and both drop a lot. Neither is bad. But the numbers you posted are not returns and you'll make a worse decision if you treat them as such.

Canton covered the main point. One more worth adding: time to income. Option A produces rent in maybe four months if quarry's timeline correction is right. Option B in eight to ten. That's six months of difference, roughly $11,000 of foregone rent on the B side, plus interest if borrowing during construction. Add it to B's cost and B goes to about $189,000 effective. Does that change the answer? It probably shouldn't, because B's rent advantage compounds for thirty years and the six months is once. But it belongs in the model.

Ledger's framing is the right one. Run both to a ten year hold with a 3% rent growth assumption and a terminal value that reflects what an appraiser will actually credit, which is halyard's question. If your appraiser says a detached permitted unit adds $120,000 of value in your market and an internal conversion adds $40,000, the whole thing is decided and you can stop building spreadsheets. The value add is the variable with the widest range and the least data, so it is worth paying to narrow it.

Reading this whole thread as somebody still figuring out where I fit, and what strikes me is that eleven months of care produced a good post. Anchor, you're not indecisive, you're missing three inputs. Fees, insurance, appraised value.

Trades side. Check your electrical service before anything. A 440 foot unit off a subpanel is usually fine, a 700 foot two bedroom detached with its own meter often means the utility sets a new service, and that's a queue you don't control. I've watched jobs sit six weeks waiting on a meter set.

Solder, does that apply if he keeps it on one meter and bills utilities as part of rent? Or does the code force separate service on a detached unit?

From a management standpoint, a 440 square foot one bedroom attached to the owner's house rents to a different tenant pool than a 700 foot two bedroom cottage. The small attached unit gets singles on short horizons and turns more often. The two bedroom detached gets couples and small families who stay two or three years. Turnover cost is real money, call it a month of rent plus $800 every time. Over ten years that difference alone is worth thousands and it never shows up in the initial spread.

Trellis, that is the most useful point in the thread, thanks. Anchor, apologies for the drift.

A permitted 700 foot two bedroom at $178,000 that rents at $1,850 is a number most markets cannot produce right now. If the ordinance permits it and the covenants do not block it, the only real hesitation on B is the cash.

How is it being funded? That is still not in the post. A HELOC at a variable rate on $178,000 during an eight month build is a very different project from $178,000 of cash. If you are borrowing, price the construction period interest and put it into canton's math. There is also a Fannie Mae policy effective March 2026 that lets projected ADU income count toward qualifying on an owner occupied purchase, capped at 30% of qualifying income. It does not apply to a house you already own. It does matter to your resale story, since a future buyer may be able to qualify partly on the unit's rent. Confirm the current terms with a lender directly rather than relying on a summary.

Tender's resale point is the one worth underlining. If buyers can finance partly on the unit's income, the pool of buyers for a house with a permitted detached unit gets wider than it is today. That is an argument for the permitted detached build over the conversion, assuming permits and appraisal both land.

Coming back to say I'd flipped to B after reading trellis and sable. Turnover and buyer pool both point the same way, and the parking issue I raised only exists in option A.

Anchor, one last practical thing. Whichever you build, get it permitted and get the final inspection card. A common pattern is a small multifamily where a previous owner added an unpermitted basement unit, and the next owner spends two years and about $19,000 legalizing it while the appraiser gives nothing for the space until the certificate is issued. Unpermitted space rents fine and appraises at zero.