How to Read Recent Comparable Sales in Your Auckland Suburb

A genuine comparable sale is a property that a buyer would have seriously considered instead of yours: sold recently enough to reflect the same market, on similar land, in similar condition, with a similar title, and competing for the same buyers. If it fails any one of those tests, it is a data point, not a comp. That distinction is the whole game. Most vendors I sit down with have not collected comparable sales at all. They have collected sale prices off a map, which is a very different and far less useful thing.

I am Amit Sharma, a licensed salesperson with Bayleys in One Tree Hill, and reading comps properly is most of what an appraisal actually is. So here is the method, in the order I use it, along with the places I most often see people go wrong.

Start with sold, not asking. An asking price is one person's opinion, usually the person who wants the highest number. A sale price is two parties agreeing in writing, with money and a solicitor behind it. Listings that are still on the market tell you about competition and about seller expectation, and both matter, but they never tell you what someone was willing to pay. If a neighbour is asking a big number and has been asking it for four months, that is evidence against the number, not for it.

Then decide what recent actually means. Recency is not a fixed number of months, it is a question of whether the market has changed since. In a flat, steady stretch, a sale from six or eight months ago can still be perfectly readable. In a turning market, a sale from three months ago can already be stale. The honest way to handle an older sale is not to adjust it with some invented percentage, it is to use it as context and weight it less. If every comp you are relying on is old, the real answer is that your suburb has not traded much lately, and you should widen the net rather than pretend the old numbers are current.

Next, match the buyer pool, not the distance on the map. This is the test people skip. Two houses can be four hundred metres apart and be sold to completely different people, because a school zone boundary runs between them, or one sits on an arterial road and one sits on a cul-de-sac, or one is a walk to the train and one is a drive. Buyers do not shop by radius. They shop by shortlist. Ask yourself the simple version of the question: would the buyer who bought that house have turned up to my open home? If the answer is no, it is not your comp.

Then check land, title and what the site allows. Floor area gets all the attention and land does most of the quiet work. A house on a flat, regular, wide site is worth something different to the same house on a steep sliver with a long shared driveway, even when the two sit side by side. Title type matters too. Freehold, cross-lease and unit title are genuinely different products with different buyer pools, different lending appetite and different resale friction, and comparing across them without saying so is one of the fastest ways to produce a misleading number. Underlying zoning matters as well, because on some sites part of what a buyer is paying for is future options rather than the house standing there today.

Now be brutally honest about condition. This is the adjustment vendors find hardest, because everyone quietly believes their own home presents better than it does. A comp with a new kitchen, a new bathroom, new joinery and a landscaped yard is not the same product as a tidy original home, even if the bedroom count and the land size match exactly. Equally, a comp that sold with obvious deferred maintenance sets a floor rather than a benchmark. The useful discipline is to walk the comps if they are still open, look at the listing photos properly, and write down what the other house had that yours does not, and what yours has that it did not.

Read the terms as well as the number. A sale price is the headline; the terms are the story behind it. An unconditional auction result and a negotiated sale with a long finance and building condition are not the same transaction, even at the same number. So are a sale with a settlement date that suited the buyer perfectly and one where the buyer paid for the privilege of a very short settlement. If your agent can tell you the method of sale, how many bidders or offers there were and whether it was conditional, you are getting a much richer picture than the price alone.

Know what the public record does not show. Chattels vary wildly between sales, and a fully chattelled home can record a higher number for the same underlying property. Related party and family transfers show up as sales but were never market tested. Mortgagee and estate sales often carry a discount for certainty and speed. Vendor incentives, rent-back arrangements and contributions to costs do not appear anywhere. None of this makes public sales data useless, it just means you should treat any single sale as a claim to be checked rather than a fact to be built on.

On sources: council records, homes.co.nz, OneRoof and the portals will all give you sold data, and your agent can pull the fuller picture including method of sale and days on market. One thing to be clear about is the rating valuation. A council capital value is a mass appraisal figure set at a common valuation date for the purpose of allocating rates. It is not a market price, it does not know your renovation, and using it as a benchmark, in either direction, will mislead you. The same goes for automated online estimates, which are useful for bracketing and poor at anything unusual. If you want the wider suburb context behind the numbers, the Ellerslie suburb data profile is the kind of background layer worth reading alongside the individual sales, rather than instead of them.

Once you have your set, sort it into three lists rather than averaging it. List one: comps clearly better than your home. List two: comps genuinely equivalent. List three: comps clearly weaker. Your realistic range sits between the top of list three and the bottom of list one, and list two tells you where inside that range you sit. This is far more honest than an average, because an average quietly hides the fact that half your comps were not really comparable. If lists one and three are crowded and list two is empty, you have learned something important: your home is unusual for the street, and the market will need to price it rather than match it.

Finish by looking at what is on the market right now. Comps tell you what buyers have already paid. Current listings tell you what those same buyers can buy instead of you this weekend. If there are five similar homes live in your suburb, your campaign has to win against them regardless of what sold in March. If there are none, you have scarcity working for you. A price built only on historic sales, with no reference to live competition, is only half a price.

The mistakes I see most often, in order: using asking prices, using the neighbour's sale from two years ago, using the rating valuation, comparing a renovated home to an unrenovated one without adjusting, ignoring title type, and quietly discarding every comp that came in lower than hoped. That last one is the expensive one. If you find yourself explaining why four of your six comps do not count, the market is telling you something and you are arguing with it.

My honest take: comps do not give you a price, they give you a range and a direction. Anyone who hands you a single confident number without showing you the sales behind it is selling you a feeling. Ask to see the comps, ask why each one was chosen, and ask which ones were left out and why. You can see the sales I have been part of on my recent sales page, and if you want a starting range for your own place before you talk to anybody, run the free AI appraisal on the home page and then check it against the comps yourself. That order, estimate first and evidence second, is the one that keeps you in control of the conversation.