Recession Budgets and the Customers You Already Have

Customer retention in a recession comes down to one decision: when the budget gets cut, protect the people who talk to your existing customers before you protect anything that chases new ones. On December 1, 2008, the National Bureau of Economic Research confirmed what most of us had been feeling for months. The US economy peaked in December 2007 and has been shrinking for a full year. Every service manager I know is being asked the same question this month, which is where to find the savings.

The easy answer is the support team. It’s a cost centre, it’s visible on the payroll report, and a cut there doesn’t show up in the quarter’s revenue the way a cut in sales does. I understand the logic. I think it’s backwards. A customer who stays through a bad year is worth more than a new one you pay to find, and the people who keep them are exactly the ones on the phones. Cut them thin enough and you’ll see it in the churn report about two quarters after the savings looked so good on paper.

Companies have lived through downturns before, and the pattern repeats. Budgets tighten, the service team shrinks, wait times grow, and the customers who were already nervous about their own money start looking at the competition. Then everyone acts surprised.

What does a thin team sound like on the phone?

You’ve probably been on the other end of it as a customer. The agent is polite and obviously rushed, the answer is half there, and you can hear the queue in the pauses. Nobody in that conversation is doing anything wrong. The person is doing the job they’ve been left with. That’s the part budget conversations tend to miss: a smaller team doesn’t give you the same service for less money. It gives you a different service, and the customer is the first to hear the difference.

I’ve never forgotten how often a customer’s anger was about the silence and not the fault. When I could tell someone what we were doing, give them a time, and say when I’d next be in touch, they’d usually stay with us through a problem I hadn’t fixed yet. I’ve kept accounts that way without fixing a thing on the first call. It costs nothing but attention. Attention is exactly what a stretched team runs out of first.

That’s why I don’t trust the headcount line on its own. Take one agent out of a team of ten and the spreadsheet says you’ve lost 10% of your capacity. On the phones it feels like more, because the work that disappears first isn’t the calls. It’s the follow-up, the notes and the two minutes of thinking that made the next call go better.

What would I protect first?

If I had to choose, I’d protect three things in this order. The first is the ability to respond to customers on the day they contact you, which usually means protecting the people on the phones and the queue. The second is callbacks and follow-ups, since those are the first promises to slip and the ones customers remember. The third is training, even in a reduced form, because it’s what makes a smaller team good instead of merely smaller.

The reason for that order is money, not sentiment. Back in 1990, Frederick Reichheld and W. Earl Sasser looked at service businesses in Zero Defections, their Harvard Business Review piece, and found that cutting customer defections by 5% lifted profits by 25% to 85%, depending on the industry. One bank’s branch system saw the 85%. Those figures are nearly twenty years old and I still haven’t seen anything that overturns them. If anything, a recession makes them sharper, because new customers get harder and more expensive to win.

Everything else gets a harder look. I’d question every report, meeting and project that doesn’t help a customer in the next quarter, and I’d pause the things that are nice to have. It’s a different way of cutting. Instead of asking what each team can spare, I’d ask what the customer would notice, and start with the things they wouldn’t. It’s slower to work out. It also tends to leave the service intact.

Where does a tight budget actually hurt service?

It hurts first in the small gaps between contacts. When a team gets smaller, the first thing to go is the time an agent used to have to finish a call properly, write good notes and take a breath before the next one. I wrote earlier this year about the drive-through order that goes wrong because the person serving you is half somewhere else. A leaner team makes that worse, because every agent is now carrying a longer queue in their head.

Callbacks are next. A ticket that sat for two days because nobody was free to update the customer can cost you the account, and you’ll never find that loss in a cost report. It shows up months later as a cancellation with no reason attached. The account manager calls it price. It wasn’t price.

Training goes after that, since a training day is a day off the phones. It’s the wrong thing to give up. A new agent hired into a recession is replacing someone who left or was let go, and the customer on the other end doesn’t care about the reason. They only hear someone who doesn’t know the answer yet. Keep the basics running even if everything else has to shrink, and keep them on the calendar, because a training session that keeps getting moved is a training session that’s been cancelled.

Does cutting the support team actually save money?

Sometimes it doesn’t, and you can check before you decide. I’d build a one-sheet workbook and call the tab “Recession cut”. Column A holds the labels and column B holds the values, one item per row, with row 1 left for the headings “Item” and “Value”. The numbers below are round examples to show the method, not any company’s real figures. Put your own in.

Rows 2 to 7 are the inputs. B2 is the number of customers, say 1,200. B3 is the average annual revenue per customer, say $6,000. B4 is this year’s churn rate, the share of customers who leave in a year, at 8%. B5 is the churn rate you’d expect after the cut, at 10%. B6 is the number of agents you’re thinking of cutting, 3. B7 is the loaded cost of one agent for a year, salary plus taxes and benefits, at $42,000. Be honest about B7: as I showed with the shrinkage rota, one agent on the payroll isn’t one seat on the phones all year.

Rows 8 to 11 do the arithmetic. B8 is the saving, =B6*B7, which multiplies agents by cost: 3 x $42,000 = $126,000. B9 is the extra customers you lose, =B2*(B5-B4). The brackets work out the rise in churn first, 10% minus 8% = 2%, and the formula then applies that to the customer base: 1,200 x 0.02 = 24 customers. B10 is the revenue those customers took with them, =B9*B3, so 24 x $6,000 = $144,000. B11 is the net result, =B8-B10, the saving minus the lost revenue: $126,000 – $144,000 = -$18,000.

A minus sign in B11 means the cut costs more than it saves, and that’s before you count the cost of winning those 24 customers back. Change B5 to 9% and the net goes positive: 1,200 x 0.01 = 12 customers, 12 x $6,000 = $72,000 lost, and $126,000 – $72,000 = $54,000 saved. That’s the point. The whole decision hangs on how much churn you believe a thinner team causes. Most budget meetings never ask that question out loud. Make them.

What do you cut instead?

Start with work the customer never sees. Reports nobody reads, meetings with no decision at the end and the third review of a process nobody has changed in a year all cost real hours. Most service operations carry more of that than the managers running them would like to admit. Find the hours there first, and put the ones you find back into the queue.

Then look at where the contacts come from. Every call that exists because of a confusing invoice, a missing status update or a broken self-service page is a call you chose to pay for. If your ticket categories are any good, the top three causes are already sitting in last month’s report. Fix them and the volume falls without anyone losing a job. It’s slower than a headcount cut. It works, because you’re removing the demand instead of the people who answer it.

If you do have to reduce the team, do it with a plan for what you stop doing, written down before the first person leaves. Tell the customers who’ll feel it, and set an honest response time rather than the old one you can no longer meet. An angry customer who was told the truth is far easier to deal with than one who was promised the usual service and didn’t get it.

How do you know retention is slipping?

Watch the quiet signals before the cancellations arrive. Contact volume per customer, repeat contacts on the same issue and the time to your first reply all move weeks ahead of churn. If you’re measuring only the customers who left, you’re reading last quarter’s newspaper.

I’d track them on a second tab called “Weekly signals”, one row per week. Column A is the week starting date. Column B is total contacts that week. Column C is repeat contacts, meaning a customer who came back about the same issue within seven days. Column D is the number of different customers who contacted you. Column E is the average hours to first reply. Two formulas do the work. F2 is the repeat rate, =C2/B2, repeats divided by all contacts: 180 / 1,500 = 0.12, or 12%. G2 is contacts per customer, =B2/D2: 1,500 / 1,000 = 1.5. Copy both down a row each week.

One week tells you nothing. Look for a run. If the repeat rate climbs from 12% to 15% over a month while first-reply hours in column E creep up, customers are coming back because the first answer didn’t stick, and some of them are starting to wonder whether you’re worth the trouble. That’s the moment to move people back onto the queue, not the moment the cancellation letters arrive.

Customers remember how they were treated when things were tight, and they remember it longer than the discount or the feature that came afterwards. A company that kept returning calls and answering honestly through a bad year earns a kind of credit that a marketing budget can’t buy. Show me a company that held onto its service team through 2008 and I’ll show you one that came out of it with customers it didn’t have to buy back.

If I were writing to my own team this month, I’d keep it short. Nobody is going to be asked to do more with less and call it a plan. We’ll be honest with customers about response times where they slip. Then I’d go back to the numbers and check that I’d actually done what I said. I’m not sure what 2009 looks like, and I don’t think anyone really is. I do know which customers I’d rather still have when it’s over.