Section one

The finding that reorganizes everything else

For thirty-eight years, the Cost vs. Value Report has asked the same question: what does a project cost, and what does it return when the house sells? The answer has been stable for most of that run, and it is not the answer most homeowners expect.

Small exterior replacements win. Large interior remodels lose. In the most recent edition, eight of the ten highest-returning projects nationally are exterior replacements — doors, siding, stone veneer, a garage door. Not one major kitchen or bathroom project appears in the top ten.

There is a plain reason for this. A buyer walking up to a house is pricing two different things at once. The exterior tells them what condition the house is in, and condition is not a matter of taste — a sagging garage door and a rusted steel door read as deferred maintenance to every buyer who sees them. The interior tells them what somebody else’s taste looks like. A $153,000 upscale kitchen is worth $153,000 to the person who chose the finishes. To the next buyer it is a kitchen, and one they may well redo.

Below is the complete regional ranking — all twenty-eight projects the report measures. These are East South Central figures — the Census division containing Alabama, Mississippi, Tennessee and Kentucky — not national averages, because regional spreads in this data are wide enough to change the ordering.

A word on what this data is and is not. The Cost vs. Value figures are estimates built from contractor cost surveys and REALTOR® valuation surveys, not from paired transactions of the same house before and after a project. They are the best long-running dataset of their kind and the industry runs on them, but they are survey-derived. Treat them as a reliable ordering of projects and an approximate magnitude — not as a promise about your house.

The five that actually pay

1. Garage door replacement — about $4,300, returns about $9,000 (211%). The best-returning project in the table, regionally and nationally alike. A garage door is frequently the single largest visual element on the front of a Huntsville ranch, and a dented, mismatched, or obviously dated one drags the whole elevation with it.

2. Steel entry door replacement — about $2,300, returns about $4,300 (186%). The cheapest project on the list and the second best-returning. Note the material: the fiberglass “grand entrance” version of the same idea costs about $10,700 and returns about 55%. Same door opening, one-third the return.

3. Backup power generator — about $11,600, returns about $16,200 (139%). This one is genuinely regional and worth pausing on. Nationally a standby generator returns about 95%. In the East South Central region it returns 139%. North Alabama buyers have priced in what spring storm season does to the grid, and they pay for a house that stays lit.

4. Manufactured stone veneer — about $10,900, returns about $13,900 (128%). Replacing vinyl skirting or a tired lower facade with adhered stone. It works because it is doing the same job as the garage door: changing what the house looks like from the street for a contained sum.

5. Minor kitchen remodel — about $27,000, returns about $28,300 (105%). The only interior project in the region that reliably returns its cost, and the definition matters enormously. The report’s “minor” kitchen means keeping the cabinet boxes and refacing them, replacing the countertop and sink, swapping the appliances for new midrange units, new flooring, paint. It does not mean moving plumbing or changing the footprint. The moment the layout changes, you are in the major-remodel tier, and the return falls by half.

That sentence is close to the whole lesson. Replacements return; additions and upgrades generally do not. The exception is the standby generator, which is a genuine addition and the only one in the table that clears its cost — and it clears it because in this region it is not really an amenity. It is insurance against a known local problem.

Section two

The upscale cliff

The Cost vs. Value data measures four projects in both a midrange and an upscale version. In all four, the upscale version returns a smaller share of a much larger number. This is the most expensive pattern in the dataset and the one homeowners fall into most often, because the decision never presents itself as “should I do a worse project.” It presents itself as “while we are in here, should we do it properly.”

The percentages understate the problem, because they are shares of very different totals. The useful way to look at it is to ask what the extra money buys. Take the midrange version as the baseline, then ask what the upgrade to upscale costs and what it returns:

• Primary suite addition. Upscale costs about $158,300 more than midrange and returns about $12,900 more. The marginal dollar returns roughly eight cents.
• Major kitchen remodel. Upscale costs about $75,200 more and returns about $14,500 more. Roughly nineteen cents on the marginal dollar.
• Bathroom addition. Upscale costs about $46,400 more and returns about $10,600 more. Roughly twenty-three cents.
• Bath remodel. Upscale costs about $52,300 more and returns about $16,100 more. Roughly thirty-one cents — the best of the four, and still a loss of sixty-nine cents on every marginal dollar.

The kitchen ladder shows the same thing in one line. In this region a minor kitchen remodel costs about $27,000 and returns 105%. A major midrange kitchen costs about $77,900 and returns 54%. A major upscale kitchen costs about $153,100 and returns 37%. Spend 5.7 times as much and recover about a third as large a share of it.

None of this is an argument against doing an upscale renovation. It is an argument for knowing which kind of decision you are making. If you are going to live with a kitchen for fifteen years, spending $153,100 on it can be an excellent use of money — it is simply not an investment, and it should not be defended as one. The mistake is spending like an investor and expecting to be repaid.

Section three

The worst choice is the one you make against your own street

Appraisers have a name for the single most reliable way to destroy money in a renovation. It is called the principle of regression, and the California Department of Real Estate’s own appraisal reference states it plainly: “The worth of a greater-valued object is reduced by association with many lesser-valued objects of the same type.”

In ordinary language: a house is valued against the houses around it, and there is a ceiling on what any street will bear no matter what you spend inside the walls. Improvements that push past that ceiling are what the trade calls a superadequacy — and the textbook example the State of California reaches for is a renovation: “a swimming pool that costs $60,000 to construct while the market is only willing to pay $10,000 for the pool.”

Fannie Mae operationalizes the same idea in the rule every appraisal on a conventional loan runs under. If a property meets market resistance because its improvements “are not compatible with the neighborhood,” the appraiser is required to address the effect on value and marketability. Your addition does not have to be ugly or badly built to be penalized. It only has to be out of step.

What the ceiling actually is, on six Huntsville streets

This is where a national report stops being useful and local closed sales take over. Below is the median price per square foot buyers actually paid in six Huntsville neighborhoods this year, against a rough figure for what quality new construction costs to build.

Read that chart as a set of ceilings. In Jones Valley the median buyer paid $184.12 a square foot. Suppose you add 400 square feet there. At $250 a foot to build — a rough figure, and one worth replacing with a real bid before you rely on it — that is $100,000. At the neighborhood’s own median rate, the market values that space at about $73,600. You would be roughly $26,000 underwater before counting the months of disruption.

And that calculation is generous, because it assumes the new square footage is valued at the same rate as the first 3,600. It usually is not. Price per square foot falls as homes get larger — a 4,000-square-foot house rarely sells for a third more than a 3,000-square-foot house on the same street. The marginal foot is worth less than the median foot, so the real recovery on that addition is probably below $73,600, not above it.

Now look at the other end of the chart. Twickenham and Old Town command $339.58 a square foot on a median home of just 1,956 square feet. That is close to double Jones Valley’s rate for houses little more than half the size. Nobody is paying for square footage in the historic districts. They are paying for the street, the age, and the architecture. An addition there buys you less than it does almost anywhere, because size was never what the price was made of.

Two honest limits on this section. First, no published study quantifies what fraction of an over-improvement a seller recovers — I looked hard, and if you see a specific percentage asserted online it is very likely invented. The principle is well established in appraisal practice; the exact penalty is not a measured number. Second, a neighborhood median is a blunt instrument. The right ceiling for your house is the ceiling on your block, which is a question about a dozen specific sales, not a neighborhood average.

Section four

Huntsville’s own evidence that conformity is priced

The regression principle is usually discussed as theory. In Five Points this year it showed up as a number, and it is one of the more striking figures in any of these reports.

Five Points is an old neighborhood — bungalows and cottages on small lots near downtown, with a median year built of 1951 among this year’s sales and seven in ten of them built before 1960. A handful of new houses have been built there on infill lots. Those new houses are, by any objective measure, better houses: larger, better insulated, under warranty, with modern systems and no knob-and-tube surprises.

Homes built in Five Points since 2015 took a median of 146 days to go under contract. Homes built before 1960 took 63.

The new construction took more than twice as long to find a buyer, in the same neighborhood, in the same market, at the same time. A buyer choosing Five Points is choosing a 1930s bungalow on a street of 1930s bungalows. A brand-new house there is competing for a buyer who, by definition, went looking somewhere else first.

That is the regression principle with a number attached, measured on Huntsville streets. It is also the clearest possible warning to anyone planning a renovation that will make their house the outlier on its block — even an outlier in the direction of “better.”

The full Five Points report has every sale behind that figure, including the days-on-market for each individual house.

Section five

Five specific ways homeowners destroy resale value

1. Adding an in-ground pool

The National Association of REALTORS® and the National Association of Landscape Professionals put a number on this in their 2023 outdoor features study: an in-ground pool cost an estimated $90,000 nationally and recovered an estimated $50,000 — 56%. More telling than the percentage is the professional judgment alongside it. Of every project in that report, the in-ground pool is the one 1% of REALTORS® recommend to a seller before listing — the lowest figure for any project in it. Basic lawn care, by comparison, recovered 217% of its cost.

Pool value is also intensely local. Redfin’s analysis of 2018 sales across twenty metros found a pool worth about $95,000 in Los Angeles and only about $11,600 in Phoenix — where a third of homes already have one. In Boston, homes with pools sold for about $15,500 less than comparable homes without — the only metro in that study where a pool was a net negative. A pool is not a fixture with a value; it is a preference with a climate and a saturation rate. (Those are 2018 sales, published in 2019 — eight years old now, and no comparable study has been run for Huntsville.)

The honest framing: a pool is a lifestyle purchase. The NAR study gave it a Joy Score of 10, the maximum on their scale. If you want one and will swim in it for a decade, buy it. Just do not tell yourself it is an investment, and know that a meaningful share of buyers will privately price it as a liability with a fence requirement.

2. Converting the garage

This is the mistake with the most specific and least understood rule behind it. Since April 2022, appraisals on loans sold to Fannie Mae must use the ANSI Z765-2021 measuring standard. Under it, a converted garage counts as gross living area — the square footage your house is valued on — only if all four of these are true:

• The required permits are in place.
• The space is finished.
• The space has conventional heating. The standard is explicit that a space heater does not qualify; forced air, radiant, ductless and solar are the examples given of systems that do.
• Ceiling height is at least seven feet (with allowances for sloped ceilings; no portion under five feet counts at all).

Miss any one and the space does not vanish from the appraisal — but it stops being square footage. Fannie Mae’s guidelines require non-conforming finished area to be reported on a separate line of the sales comparison grid with its own market adjustment. It gets valued as an amenity rather than at your home’s per-square-foot rate, which is very nearly always a worse outcome.

Then there is what you gave up. Redfin’s analysis of 2018 sales found homes with garages sold for about $23,200 more than comparable homes without — a 12% premium nationally. And the NAHB’s survey of more than 3,000 buyers found 81% rate garage storage essential or desirable. Be precise about what that evidence is: it compares homes with and without garages, which is not the same as a study of conversions. No study of conversions specifically exists. But the direction is not in doubt.

3. Doing the work without permits

Here the mechanism is well documented and the magnitude is not, so I will give you the mechanism and decline to invent the rest.

Fannie Mae’s Selling Guide requires that where an appraiser finds an addition without the required permit, they “must comment on the quality and appearance of the work and its impact, if any, on the market value.” Note the phrase if any — the rule is not that unpermitted work is worthless. It is that the work gets judged on its merits by someone with no reason to be charitable.

The sharper consequence is the one from the section above: unpermitted finished space fails the ANSI test and stops counting as gross living area. A homeowner who finishes 500 square feet without a permit often believes they have added 500 square feet to their house. On the appraisal, they have not. They have added an amenity.

The rules get harder still for an accessory dwelling unit that violates zoning: Fannie Mae requires the appraisal to demonstrate acceptability using at least two comparable sales with the same non-compliant zoning use, within a minimum of three settled sales. In most Huntsville neighborhoods, finding two such sales is not a paperwork problem. It is an impossibility, and that is how a financing contingency dies.

I could not find a single credible study quantifying the sale-price discount on unpermitted work — not from NAR, the appraisal journals, Zillow, Redfin, or the academic literature. Articles asserting a specific percentage are not citing research. What is documented is the process, and the process is enough to take seriously.

4. Buying the solar system and expecting it back

Solar is the third-worst-returning project in the regional table — only an upscale primary suite addition and a backyard patio do worse: about $47,800 spent, about $16,900 returned — 35%. The reason is well established in research that deserves to be better known.

Fannie Mae and Lawrence Berkeley National Laboratory studied 554,947 transactions across four states through early 2020. Homeowner-owned systems carried a real premium: +1.63%, about $14,500 on an average 5.5 kW system. But the premium had collapsed over the decade — from roughly $6 per watt in 2011 to under $0.40 per watt after 2016, a fall of more than 75%, converging on the value of the energy the panels actually produce.

And the common belief about leases is wrong in an interesting way. Berkeley Lab’s dedicated study of third-party-owned systems found no statistically significant premium and no discount — and no measurable difference in days on market. A lease is not a price penalty. It is friction: the review found UCC filings on 77% of the leases examined, which your buyer’s lender will find at closing and will want resolved. Plan for the paperwork, not for a haircut.

5. Painting the house in colors the market prices down

Zillow has run this study three times, and the two designs need to be told apart or the numbers contradict each other.

Its 2017 work was a hedonic analysis of more than 32,000 photos from homes that actually sold between 2010 and 2017 — revealed behavior, controlled for size, age, location and date. It found a white or colorless bathroom sold for about $4,000 below expectation, and a red dining room about $2,000 below.

Its 2025 and 2026 studies are a different instrument: survey experiments in which 4,200–4,400 buyers were shown rooms in randomly assigned colors and asked what they would pay. The 2026 edition found the largest penalties yet — a fire-hydrant red bathroom about $8,000 below expectation, an ochre yellow kitchen about $6,600 below, and ochre yellow applied across all four tested rooms costing an estimated $18,164.

Stated willingness to pay overstates real price effects, and the two designs disagree about which colors win — so use the survey numbers as a ranking of risk, not a price list. What survives both: saturated, unusual color in a wet room is the most expensive paint decision available to you, and it costs a few hundred dollars to undo.

This last point is where most advice on this subject goes wrong. Zillow also analyzed more than two million listings against 600 listing phrases and found that homes described as having “custom” features sold for 3.2% above expectation, and “bespoke” finishes 3.0% above. Quartzite counters, outdoor kitchens and gourmet kitchens all carried premiums. Buyers are not punishing individuality. They are punishing a red bathroom.

Section six

The unglamorous things that are worth more than they look

Two categories of work are systematically undervalued by homeowners, because neither one photographs well and neither feels like an improvement. Both are among the best money a seller spends.

Condition, which buyers price ruthlessly

Zillow’s analysis of more than two million 2024 listings found that homes whose descriptions read as a fixer-upper sold 7.3% below expectation — the widest such discount in three years. Listings using “needs work” or “TLC” ran near 8% below, which on a typical Huntsville sale is tens of thousands of dollars. Remodeled homes sold 3.7% above.

The gap between those two numbers is roughly eleven points of sale price, and most of what separates them is not a renovation at all. It is a roof that does not have three cracked tabs, a water heater without a rust ring, gutters that are attached, and a crawlspace that does not smell. Deferred maintenance is not priced as the cost to fix it. It is priced as the buyer’s worst guess about what else is wrong, and that guess is always more expensive than the repair.

Floors, and the interior work that returns the most

In NAR’s 2022 Remodeling Impact Report, refinishing existing hardwood floors recovered 147% of its cost and installing new wood flooring recovered 118% — both national figures, and both higher than any interior project in the regional Cost vs. Value table. The NAHB’s buyer survey puts hardwood on the main level among the features 81% of buyers call essential or desirable.

Which makes carpeting over sound hardwood one of the quietly worst decisions available in an older Huntsville home — and there are a great many mid-century houses inside the loop with oak under the carpet. To be careful about the evidence: no study prices “carpet installed over hardwood” directly. This is an inference from two solid datasets, not a measured effect. But it is an inference in one direction only.

Section seven

What to do in the ninety days before you list

If the question is narrowly “what should I spend money on before selling,” the research converges on a short and rather boring list. Notice how little of it is renovation.

NAR asked REALTORS® which projects they actually recommend to a seller before listing. The answers, in order: paint the whole interior (50%), paint one room (41%), new roofing (37%), a kitchen upgrade (30%), a bathroom renovation (24%). An in-ground pool: 1%.

• Paint, in ordinary colors. The cheapest lever with the best evidence behind it, and the fastest way to undo a color decision the market prices down.
• Fix the front door and the garage door. The two highest-returning projects in the entire report are both on your front elevation and together cost under $7,000 in this region.
• Deal with the roof honestly. An asphalt replacement returns about 81% regionally, which is a loss — but a roof at end of life is not a renovation decision, it is a negotiation you will have anyway, on the buyer’s terms with the buyer’s contractor’s estimate.
• Refinish the hardwood, do not cover it.
• Clear the deferred maintenance list before an inspector writes it for you. Gutters, grading, crawlspace, the water heater, the two outlets that do not work.
• Do not start a major kitchen remodel or an upscale bath to sell. A midrange major kitchen returns 54% regionally: you will spend $77,900 to move the price roughly $41,800, and delay the listing by months. A midrange bath remodel is a different proposition at 85%, and a cosmetic refresh of one is better still.

One caveat that matters more than any figure above: a project that returns 60% of its cost can still be the right decision if the alternative is not selling at all. A house that will not attract an offer in its current condition has an effective return of zero on everything. These percentages describe the marginal dollar on a saleable house — not the first dollar on an unsaleable one.

Section eight

What this data does not say

A great deal of renovation advice online is confidently wrong, and the most useful thing I can do is mark the places where the evidence runs out. Each of these is a claim I went looking for and could not support.

• “Removing a bedroom costs you X%.” No such figure exists, and the academic literature actively complicates the premise. In a review of the hedonic house-price literature, bedroom count appeared in forty studies: it was positive in 21, negative in 9, and statistically insignificant in 10. Bathroom count, by contrast, was positive in 34 of the 40 studies it appeared in. Combining two small bedrooms into one large one is a real risk to your buyer pool, but anyone quoting you a percentage is guessing.
• “Unpermitted work reduces your sale price by X%.” No study anywhere. The process consequences are documented; the price effect is not.
• “Wallpaper / bold tile / themed rooms cost you money.” Plausible, universally asserted, and supported by no price data from any credible source. Zillow has studied paint colors, not wallpaper.
• “A house with no bathtub is hard to sell.” The most-cited source for this traces back to a survey commissioned by a bathtub manufacturer, and it contains no resale data at all.
• “You recover X% of an over-improvement.” The appraisal principle is solid. The percentage is invented every time you see it — including, to be fair to the point, the illustrative arithmetic in Section three, which is a worked example on one set of assumptions rather than a measured recovery rate.
• “A pool hurts you below a certain price point.” No study segments pool value by price tier. It is a reasonable intuition with nothing behind it.

I would rather hand you six honest gaps than six confident numbers I cannot defend. If a renovation decision turns on one of these questions, the answer is not on the internet — it is in the last dozen sales on your street.

Section nine

Find your own ceiling

Every figure in Section three comes from closed MLS sales, neighborhood by neighborhood. Each report below lists every sale behind its numbers, including size, year built, days on market and price per square foot — which is what you need to work out what your own street will bear.

• Jones Valley — The lowest price per square foot in this series at $184.12 — and therefore the tightest ceiling on any addition.
• Blossomwood — Mid-century houses on big lots at the foot of Monte Sano. A median 14.5 days to contract, the second fastest of the six.
• Piedmont — The fastest contracts of the six, a median 6.5 days — and only two of ten sellers got their asking price.
• Five Points — Where new construction took 146 days to go under contract against 63 for the pre-1960 houses around it.
• Twickenham & Old Town — $339.58 a square foot on a median home of 1,956 square feet. Nobody there is paying for size.
• The Ledges — The highest price per square foot measured, $370.55, in Huntsville's gated golf community.

Questions people ask

Renovation and resale questions, answered