A Data-Driven Pricing Framework for B2B Technical Services

Most B2B technical-service firms in Mexico set prices by habit: cost-plus, or last year’s number plus a bit. That is simple and defensible, but it leaves the most sensitive lever on profit unmanaged, and it can be replaced with a five-step method built from data the firm already has. This is a process, not a promised margin outcome. What follows is the method and, just as important, the assumptions each step carries.

Why price is the lever worth the work

The case for spending time on price is old and consistent. Michael Marn and Robert Rosiello argued in Harvard Business Review (September-October 1992) that the right price can lift profit faster than higher volume can, and the wrong price can shrink it just as fast. The often-quoted numbers on exactly how much are where care is needed. A McKinsey Global 1200 analysis reported by Forbes in March 2010 found that a 1% price increase, holding demand constant, raised average operating profit by about 11%.

That “holding demand constant” clause is load-bearing, and it rarely holds in full. John Dawes of the Ehrenberg-Bass Institute for Marketing Science noted in September 2018 that the double-digit framing assumes zero volume loss, while average brand price elasticity sits around -2.5, meaning a price rise usually costs some volume. Read the 11% as the ceiling of a favorable case, not a forecast for your firm. The honest takeaway is the qualitative one: price is a high-leverage lever, so it deserves a method rather than a habit.

There is also evidence that most firms know this and still under-manage it. A Bain & Company survey of more than 1,700 B2B companies, published in 2018, found that about 85% of respondents believed their own pricing decisions could improve, and that firms combining tailored pricing with aligned incentives and the right tools were far more likely to be top performers (78% versus 18%). The survey is dated, so treat it as a large, still-cited finding rather than a reading of today’s market.

The five-step framework

Run these in order. Each step names the data it uses and what it can and cannot tell you. None of it requires enterprise pricing software.

1. Establish the cost floor

Start with the firm’s own job-costing data: labor hours by role, materials and subcontracted work, and allocated overhead. From that, compute a true minimum price per engagement type - the level below which a project destroys value. This floor is specific to your accounting; no industry benchmark substitutes for it, and none is invented here. Before you commit to any structure above the floor, model what a price change does to a single engagement’s economics with the ROI calculator.

2. Map the value drivers

The floor tells you where you cannot go. Value drivers tell you where you can. Pull them from your win/loss notes and a short competitor list built from public pricing pages, RFP responses, and sales-team notes. For technical firms the drivers are usually specific: uptime and SLA commitments, certifications and compliance, response time, integration depth, bilingual support. Utpal Dholakia observed in Harvard Business Review (August 2016) that value-based pricing is the most discussed and most misunderstood pricing concept, and that misreading it pushes firms back to cost-based pricing that leaves money on the table. The point of this step is to name what clients actually pay more for, separate from what it costs you to deliver.

3. Probe willingness to pay

Guessing what clients will pay is where cost-plus firms usually stop. There is a low-cost survey method instead. The Van Westendorp Price Sensitivity Meter, introduced by Peter van Westendorp at the 1976 ESOMAR congress, asks four questions about a given offer: at what price is it too expensive to consider, at what price is it so cheap you doubt the quality, at what price is it getting expensive but still worth considering, and at what price is it a bargain. B2B International notes that as few as about 50 responses can be workable for a well-defined B2B segment. The output is a defensible price range per segment, not a single number.

State the limitation plainly: the method reads price perception, not demand. By design it measures how buyers judge a price, not how price affects purchase volume - van Westendorp himself framed it as a pricing tool and did not attempt to solve demand estimation. It is a way to stop guessing, not a substitute for elasticity testing.

4. Build structured tiers

With a floor, value drivers, and a willingness-to-pay range in hand, design 2 to 4 tiers rather than one rate for everyone. Groups of B2B clients differ systematically in what they will pay and in what they cost to serve, so tiers should be built from your own account and segment data. The Umbrex B2B Pricing Playbook describes a widely used consulting sequence for this: clarify objectives, define three to six segments, analyze current price and margin by segment, design tier concepts, set explicit price fences (the criteria that qualify an account for a tier), pilot, then embed. A useful mechanism sits underneath: Stefan Michel argued in Harvard Business Review (March 2015) that letting customers reveal their own segment by choosing a tier can capture value that imposed segmentation misses. Build the fences from your data; let clients self-select within them.

5. Review annually against INPC and FX

Prices drift out of date. Set a fixed annual review, and drive it with two public inputs plus your own realized costs. The first input is Mexican consumer inflation. Headline annual INPC was 3.37% in June 2026 according to INEGI (published July 9, 2026 and reported by Milenio), its lowest reading since December 2020, with core inflation at 4.03%. The index itself is published monthly by Banco de Mexico, standing at 145.131 in June 2026 on the base where the second half of July 2018 equals 100.

Do not apply INPC as a blanket “raise prices by the inflation rate” rule. It is a basket average, not your firm’s cost basket, so use it as context alongside your actual cost changes. Banco de Mexico’s inflation target is 3.0% plus or minus one percentage point, as restated in Rio Times coverage of the June 2026 data, which means a 2% to 4% print is closer to normal variation than to a signal that your floor has moved.

The second input matters only if your inputs are dollar-denominated: software licenses, imported equipment, or contractor rates billed in USD. For those firms, watch the Banco de Mexico FIX rate. In mid-July 2026 the peso traded around 17.4 per dollar, and over the prior 30 days it ranged between about 17.17 and 17.63, a swing of roughly 2.7% (Rio Times, July 17, 2026). That illustrates how much the rate can move inside a few weeks; it is a reason to check exposure at review time, not a forecast of direction.

What this method will not do

The framework is a discipline, not a guarantee. It will not deliver a specific margin or revenue percentage. The Bain and McKinsey-lineage figures above describe large, mixed samples under a stated no-volume-loss assumption, not an outcome any single firm should expect. It will not hand you a Mexican technical-services rate benchmark, because no reliable one exists to cite and none should be invented; your floor and your willingness-to-pay range come from your own data. And it does not touch tax or legal ground. IVA and CFDI facturacion are real considerations when you change prices, but route those through your own accountant or fiscal advisor rather than any general guidance.

Worth repeating, because an industry survey suggests the gap persists: Simon-Kucher’s Global Pricing Study 2025, a survey of more than 2,200 business leaders across 28 countries and 39 industries, reported that companies realize less than half of their intended price increases on average, and that 54% of firms not adopting AI cited a lack of in-house expertise. The method here is aimed squarely at that gap. It does not need new tools or a research budget. It needs the firm to use the cost, win/loss, and account data it already keeps, add one small survey, and review the result on a schedule against inflation and the exchange rate.