Choose a price. Test the response.

Before choosing a number, define the offer, the customer and the unit they pay for. A software licence per employee, a subscription per factory and a fee per completed transaction create different incentives and revenue patterns.

Also establish whether you are pricing a new product, repricing existing customers or responding to a competitor. Raising an existing price creates migration and retention questions. A new offer requires evidence about adoption and a comparison with what buyers do today.

Suppose an industrial software company wants to price a predictive-maintenance service. It alerts a plant before equipment fails. The client wants a profitable launch, and the initial target is manufacturers with costly unplanned downtime.

The useful first question is: which plants obtain enough value to justify buying and implementing it?

Assume a target plant loses £5,000 of contribution for each hour of unplanned downtime. The service is expected to prevent ten hours per year, creating £50,000 of annual benefit. Implementation and additional monitoring work cost the customer £10,000 in the first year.

That leaves £40,000 of first-year benefit before the software fee. At a £20,000 annual price, the plant would retain £20,000 of that benefit in year one under the stated assumptions.

Use lost contribution, rather than simply multiplying hours by sales. A stoppage may delay production that can be recovered later; some operating costs may also be avoided. Ask what the downtime figure represents before treating it as economic value.

The calculation is a way to understand the offer, not permission to charge £40,000. The customer may doubt the claimed improvement, prefer a cheaper preventive-maintenance option or require a substantial return for taking implementation risk.

For a second plant where downtime costs only £1,000 an hour, the same ten hours avoided create £10,000 of gross annual benefit. A uniform £20,000 price would be difficult to justify from this benefit alone. That suggests a narrower initial segment, a different package or a different commercial model.

Ask what happens if the plant does not buy. It might continue with its existing maintenance schedule, add inspections, purchase a competing service or replace an unreliable machine.

Compare the outcomes and total costs of those options. A competitor's headline fee is incomplete if it excludes installation, requires extra staff or offers a different level of support.

This also helps explain your own price. “We are 10% cheaper than the competitor” is a position. “Our service gives this type of plant a better expected result after implementation costs” is a value proposition you can test.

The buyer's budget and decision process matter too. A plant manager may understand the operational benefit while a central procurement team controls the contract. Identify who must be persuaded and what evidence they need.

Suppose the service costs £5,000 per plant per year in hosting, monitoring and variable support. At a £20,000 fee, annual contribution is £15,000 per plant before acquisition costs and fixed development and operating costs.

If annual fixed costs are £600,000, 40 fully paying plants cover those costs at that contribution. This is a simplified steady-state break-even calculation. A launch with customers joining throughout the year, onboarding expenses and upfront development spending needs a cash and timing model as well.

Include the costs caused by the proposed offer. An attractive software margin can disappear if each customer needs extensive custom integration or high-touch support. Connect the price and package to what the business must deliver.

Do not use an allocated accounting cost as an automatic minimum price for every short-term decision. Equally, a price above variable cost does not establish a sustainable business: someone still has to pay for the fixed capability and investment.

A fixed subscription per plant is simple to budget and administer. Charging per monitored machine may fit plants of different sizes, provided machine count is a reasonable guide to value and service cost.

An outcome-linked fee sounds appealing because the customer pays for results. The difficulty is agreeing how to measure downtime avoided. You cannot directly observe the breakdown that did not happen, and the plant's maintenance behaviour also affects the result.

Choose that model only when both sides can agree on a credible baseline, measurement process and allocation of risk. Otherwise the commercial dispute may cost more than the apparent alignment creates.

You can also use packages: core alerts, more detailed analytics or additional support. Make the differences useful to identifiable customer groups. Arbitrary complexity makes the sales discussion harder without necessarily improving the economics.

Now consider a separate repricing example. A business charges £100 per customer per period, with £40 variable cost. It proposes increasing the price to £110, keeping the service and variable cost unchanged.

To preserve total contribution, the business must retain £60 ÷ £70, or about 85.7%, of its original volume. It can lose about 14.3% before contribution falls, assuming fixed costs and customer mix stay the same.

With 1,000 original customers, current contribution is £60,000. Retaining 900 produces £63,000 at the new price. Retaining 800 produces £56,000.

That threshold is more informative than observing that the new margin is higher. It tells you what you need to learn about the customer response. If the customers who leave are disproportionately expensive to serve, model their contribution separately instead of assuming identical customers.

For a new product, use customer discussions to understand value and alternatives, then seek evidence of willingness to buy at the proposed terms. Enthusiasm for a concept is weaker evidence than a purchasing commitment.

For repricing, examine comparable past changes and use a controlled rollout where appropriate. Track realised price after discounts, retained volume, mix, support costs and contribution over a period long enough to observe renewal behaviour. Avoid calling the test a success simply because invoiced revenue rises in the first week.

Your recommendation should state the price or range, the customer segment, the pricing unit and the main uncertainty. For the maintenance service, that could be a £20,000 per-plant subscription tested first with plants whose avoidable downtime supports a compelling return, alongside explicit limits on custom support.

To practise the opening, try Industrial Hardware Pricing Pressure. Explain what customers would do differently at the proposed price and how that response changes the client's profit. If the trigger is a rival's discount, continue with the competitive-response framework.