Am I undercharging?

Reference guide from Amplifai — the structured AI workspace for NZ business decisions.

Decision · Making the money make sense

There are two ways to think about pricing. Most new businesses use cost-plus — work out what something costs you, add a margin, that's your price. It's the only method available when you're starting out, because the alternative — pricing to what customers will actually pay — assumes a customer base to read.

Once you have one, that alternative becomes available. The information starts showing up in your day-to-day: what customers react to, what they ask about, who pushes back on price and who doesn't, who comes back and who doesn't. Most operators never make the switch. They keep using the first method long after the second is available, partly out of habit and partly because asking the question — could I charge more? — feels like staking the customer relationship on the answer.

Underneath all that: pricing isn't a launch decision. It's an operating discipline that needs a review rhythm, the same way your accounts do.

The trap most operators are in

Cost-plus has one quality that makes it stick: it doesn't require you to make a value claim about your work. The number falls out of a calculation. You're not standing in front of a customer asserting this is what my work is worth — you're handing them an arithmetic result. That's psychologically much cheaper than the alternative.

The other thing that makes it stick: the price set at launch usually stays set. Cost-plus made sense when you didn't have customers; now you do, and the price has been working, so nothing forces a review. The information about what customers might pay accumulates around you but doesn't get read, because reading it requires deliberate work nobody scheduled.

The trap is mechanical, not psychological. The method you used at launch was right for an information state you no longer occupy. Your customers have been generating data about willingness-to-pay since you opened, and most of it has been going un-read.

What's already in front of you

You don't need formal price research to start reading customer signal. The signal is already in your day-to-day; you're just not always reading it as pricing information.

Discount requests. Every time a customer asks for a discount and you say no, you learn something. If they buy anyway, your price was below the point where they'd walk. If they walk, your price was at or above it. Track this across thirty customers and you'll know more about your pricing than any consultant could tell you. Most operators handle each discount request as a one-off without noticing the cumulative information.

Lost quotes. When you lose a job, the reason usually gets softened. "We went with someone else" — could be price, could be timing, could be that the other quote felt more competent. Ring them and ask, casually, a week later. "Too expensive" tells you one thing. "They could start sooner" tells you something different.

Early price questions. When a customer asks "how much" before you've explained what they're getting, they're price-shopping. When they ask after you've explained, they're value-considering. The mix you mostly get tells you whether you're attracting the buyers you want — and whether your price is in the band that filters for value-led buyers or attracts price-led ones.

Repeat business and referrals. Customers who come back without bargaining are telling you the price was acceptable. Customers who refer you are telling you the value was high enough to risk their own reputation on. If you've got both at your current prices, you're probably below the ceiling the market would pay.

Comparison-shopping mentions. When customers tell you they've been quoted by someone else, that's market-price information arriving free. Note the numbers when they come up. After ten or twenty such mentions you'll have a rough read on where competitors sit — and whether you're materially below the band, in it, or above it.

None of this is sophisticated. It's reading the situations you're already in. The discipline isn't the data-gathering; it's the deciding-to-read.

When you want better information

Sometimes the passive signals don't quite settle the question. You suspect you're undercharging but the discount requests and lost quotes haven't given you enough to act on. That's when a deliberate test earns its place.

Quote your next ten jobs slightly higher. Not double. Just enough to notice — maybe 10% above where you'd normally land. Watch what happens. If your win rate stays roughly the same, your old price was leaving margin on the table. If it drops materially, you've found the current ceiling. Either way, you've learned something specific to your business at this moment, with your customers.

Add a higher-priced option alongside your existing offer. A premium tier. An expedited service. A bundle with extras. The point isn't to replace the existing offer — it's to introduce a higher price point and see who picks it. Customers who choose it are telling you there's willingness-to-pay above where you're sitting. Customers who don't aren't lost — they still have the original offer.

Limited-quantity premium versions. If a higher tier doesn't fit your offer shape, try "we have capacity for [n] of [premium version] this month." Reveals demand at the higher price without committing it as a standing rate.

A few honest notes on what these tests will and won't tell you. What customers say they'd pay is unreliable — they consistently underestimate in survey contexts. What they actually do when prices change is the data you can trust. Build your read on what they do, not what they say. A single quote that wins or loses doesn't tell you much; the pattern across ten or twenty does. And tests need a defined window with an exit criterion before they start — "for the next ten quotes I'll price at X. If win rate drops by more than [n]%, I revert" — otherwise the test becomes the new normal by default.

Acting on what you find

If the tests confirm there's room to charge more, the next question is how to change prices without burning relationships. The fear that blocks pricing reviews is real, but most of it can be addressed by how you make the change.

New customers first. The simplest move: apply the new price to new customers only. Existing customers stay where they are. You don't have to announce anything to anyone existing. If the new price holds, it becomes your standard rate as your customer base turns over. If it doesn't, your next quote goes back, and nobody existing ever knew you tested.

Effective from a future date. For service businesses with ongoing clients, "new rates from [date]" gives you both signal-time and walk-back-time. Severe reactions in the lead-up means you have time to recalibrate before the date hits.

Anchor an increase to something external. If you have to raise prices on existing customers, anchor it to something outside your value claim — input costs went up, wages went up, insurance is up. Customers absorb cost-pass-through differently from value-based increases, even when the dollar amount is the same. You're not claiming the work is worth more; you're passing through a cost they can verify exists.

Treat key customers as deliberate exceptions. Some long-term customers warrant being held on existing rates as a retention decision. New price applies generally; specific relationships get hand-managed.

None of these asks you to bet the customer base on a price change. They all give you ways to test without staking, and ways to walk back if the result is worse than expected.

The review rhythm

Cost-plus persists past its useful life partly because nobody scheduled the review. Operating disciplines that aren't scheduled don't run.

A workable cadence:

Continuously, read the passive signals. This doesn't need to be a project — it needs to be a habit of noticing.

Annually, review prices deliberately. Look at the year's discount-request pattern, lost quotes, repeat business. Check what's happened to your costs and your competitors'. Decide whether prices need to move and by how much.

Triggered by events, review out of cycle: cost shocks, competitive shifts, capacity changes — or whenever you've gone more than eighteen months without thinking about pricing at all.

The annual review is the spine. Most operators don't do it. The exposure compounds: prices stay anchored to last year's cost structure, last year's market position, last year's customer mix. Costs creep up; prices don't. Customer composition shifts toward price-sensitive buyers because they're the ones your low price attracts, which makes raising prices harder, which keeps the cycle running.

The cadence is the work. The price itself is the result of the work.

Where this stops

Complex pricing strategy — dynamic pricing, segmentation, channel-specific or network-effects pricing — needs a CFO or strategy consultant. Different tooling, different reasoning.

Formal market research — willingness-to-pay studies, conjoint analysis, large-sample customer interviews — needs a market research firm. Useful when the stakes warrant the spend; rarely the right next move for most SMBs.

Pricing tactics that wouldn't pass the if-your-mother-knew test. Charging different customers different prices for the same offer when there's no transparent reason for the difference. AI or algorithm-driven personalised pricing where customers don't know they're being individually priced. Pricing below cost to drive a competitor out. Any of these takes the question out of am I undercharging territory and into can I actually do this thing territory. See Fair Trading Act essentials for the boundaries on differential and predatory pricing.

New-customer-vs-existing-customer pricing — the retract-on-no method above — passes the test, because the difference is transparent and the customer can see they're getting a new-customer rate. The test isn't about whether prices differ; it's about whether the difference is one a reasonable person would think was fair.

The shorter version

You probably are undercharging — most operators using cost-plus past launch are. The information that would tell you for sure is in your existing customer interactions; the work is reading it. The fear of testing higher prices is real but mostly mis-calibrated, and the methods that let you test without burning bridges exist. The actual discipline is running this on a rhythm: continuously reading passive signal, annually reviewing deliberately, and out-of-cycle when events trigger it. The cadence is the work.

Last verified 19 May 2026 against business.govt.nz pricing guidance, Fair Trading Act 1986 framework at legislation.govt.nz, and Commerce Commission Fair Trading Act operational guidance for the scope-stop on differential and predatory pricing. Full source list: references.