The cost of getting it wrong I watched a client raise prices on their core service by 18% last year. No testing. No customer conversations beforehand. Just a decision made in a spreadsheet and pushed live on a Monday morning. By Friday, they'd lost three major accounts. One customer of four years simply didn't renew. The revenue gain from the increase was wiped out within a month, and they spent the next quarter trying to win those clients back. They never did. This happens more than you'd think. And what's worse is how preventable it is. Why we skip the testing phase I get why people don't test price changes. Time pressure is real. You're running tight margins. You see the gap in your numbers and you think "we need more revenue, let's just do this." The logic feels sound: higher prices equals more money. Done. But price isn't just a number. It's a signal. It's how your customers understand their relationship with you. Mess with it without understanding the ripple effects, and you're gambling with people's trust. I've also seen what happens when price rises are implemented badly because someone else decided them. Vodafone's situation with franchisees being fined for infractions they arguably didn't commit, or the carmakers now scrambling to cover a £3bn shortfall because their loan redress calculations were wildly off target. These aren't small failures. They're the result of decisions made without proper validation, and the cost is enormous. What testing actually looks like Testing a price increase doesn't mean running a 12-week pilot. I'm talking about real conversations with real customers before you commit. Start with your best customers. The ones who pay on time, don't churn, and actually value what you do. Tell them what you're thinking. Not "would you leave if we raised prices by 15%" because that's a leading question and they'll say yes. Instead, explain why you're considering it. Talk through the value they're getting. Ask them what would genuinely break their decision to stay. You'll learn two things: whether they'll actually leave, and more importantly, what you don't understand about your own pricing. Sometimes people stay because they're locked in, not because they value you. That's useful to know before you raise prices. Next, test with a small segment. A cohort of new customers, or a specific geography, or one product line. Not everything at once. Raise prices for one group and watch what happens over 4 to 6 weeks. Track churn, track customer feedback, track the actual revenue impact. Not the theoretical impact from your spreadsheet. The real impact. Then talk to the people who leave. Why did they go? Was it the price, or was something else already wrong? Did they find a competitor, or did they cut the service entirely? Those conversations are worth more than any survey. The numbers that matter Here's what I see go wrong most often: people focus on the wrong metrics. They calculate that if they keep 95% of customers at the new price, they make more money. Mathematically true. But what if they actually keep 78% because they didn't test first? Now they've lost revenue and damaged relationships they spent years building. I worked with a small team running subscription software. They wanted to raise their annual plan from £4,200 to £5,100. A 21% increase. They tested it with 40 customers across different segments before making it live. Turns out, enterprise customers barely flinched. Mid-market customers were sensitive. Smaller accounts started looking elsewhere. So they didn't raise prices uniformly. They tiered it: 12% for the largest accounts, 18% for mid-market, and offered the smaller customers a path to a lower-tier product instead. The result was that they retained 91% of revenue, added 14% through new pricing, and didn't lose a single strategic relationship. That 14% figure came from testing, not guessing. What you're actually testing You're not testing whether people will pay more. You're testing whether your value proposition holds up under scrutiny. You're testing whether you've built something defensible or just something people tolerate. That matters because it changes how you approach it. If you discover through testing that price sensitivity is high, that's not a reason to abandon the increase. It's a reason to understand what you need to improve first. Better service? Clearer ROI? Different packaging? These are the conversations that happen before prices go up, not after. What to do this week If you're thinking about a price increase, don't implement it yet. Pick your five best customers this week. The ones who actually matter to your business. Write down three honest questions about how they perceive your value and whether they'd stay if prices moved. Schedule calls with them. Not sales calls. Just conversations. Second, if you've already raised prices recently without testing, get feedback. Not a survey. Real conversations with the people who left and the people who stayed. Find out what you missed. This data is gold for your next decision. Third, if you use a decision matrix or structured analysis tool (ALIRA has one that works well for this), map out the actual impact scenarios before you move. What's the best case? What's the realistic case? What's the catastrophic case? Then test the assumptions that sit underneath each one.