How Do You Measure Direct-Mail ROI?

Give the card its own way in, count everything that comes through it for six weeks, and set that against what a customer is worth to you. That is the whole method. The arithmetic is simple enough to do on the back of the card; the discipline of counting every call for six weeks is the part most businesses skip, and it is the part that decides whether you learn anything.

ROI, return on investment, means what the mailing brought in against what it cost. You will never get a perfect number for a mailing, because some customers will not remember where they heard of you. You can get a number good enough to decide whether to mail again, and that is the standard this guide works to.

What should you count?

Count every way a household can respond to the card, not just the phone. For most local businesses that is some mix of:

• Calls, including the ones that went to voicemail and the ones that did not turn into a job.
• QR scans, which a QR generator or your landing page can count for you.
• Landing-page visits to a page only the card points at.
• Walk-ins who mention the card, or bring it in.
• Bookingsmade online through the card’s link.
• Redeemed codes or offers that appear nowhere else.

Which of those matters depends on the business. A restaurant in West Ashley counts cards handed over at the till. An HVAC company counts calls and the service visits that came from them. A dentist counts new patients who booked. Count contacts and customers separately, because a call that did not become a job still tells you the card was seen, and a job tells you it paid.

How do you make the card countable?

You make a card countable by giving it a way in that nothing else in your marketing uses, so that anyone who arrives that way can only have come from the card. There are four ways to do it, and you want at least two.

1. A QR code to a landing page unique to the card.A landing page is a single page built for one purpose, with the offer from the card on it and one way to act. If the card is the only thing that links to it, every visit is the card’s.
2. A unique offer code. A code or a named offer that is printed on the card and nowhere else. When it comes in, by phone or at the counter, you know. This works for restaurants, salons and shops especially well, because the card comes with the customer.
3. Ask every caller, and write it down. A tally sheet by the phone, and everyone who answers uses it. This is the method that fails most often, because it depends on people rather than on technology, and it is also the one that catches customers the other methods miss.
4. A tracking phone number, if you want one. A separate number printed on the card that forwards to your main line and logs every call. The tradeoff is a monthly fee, and a number on the card that does not match your van or your Google listing. Worth it when call volume is high enough that nobody will keep the tally. Unnecessary for a business that can ask.

Use more than one because each one misses somebody. The person who scans the code will not mention the card when they call. The person who calls will not have scanned anything. The overlap is small and the gaps are real.

When should you judge a mailing?

Judge a mailing at six weeks after the drop, not two. Cards arrive over a few days, the first response comes in the first week or so, and then the tail begins: the card that sat on the fridge until the water heater went, the plumber’s card that was kept in a drawer for exactly the reason it was kept. A business that decides at two weeks has measured the first half.

Two habits make the judgment fair. Know the date the card actually landed, which the mailing calendar gives you, and compare the six weeks after it with the six weeks before, so that a seasonal lift or a slow month shows up as what it is. Keep counting after six weeks if calls keep coming, but treat those as the bonus, not the verdict.

The arithmetic, worked backwards from a customer

Start from what a customer is worth to you over a year, not from a response rate. Take the average sale, multiply it by how many times an average customer comes back in a year, and that is the value of one new customer. Divide the price of the spot by that value and you have the number of customers the mailing needs to break even.

A small spot on a shared card to 5,000 homes is $249. If one customer is worth more than $249to you over a year, the mailing needs one customer to pay for itself, and everything after that is return. If a customer is worth about half that, it needs two. If a customer is worth a tenth of it, it needs ten, which is where a restaurant’s repeat visits start to matter more than a roofer’s single job.

Notice there is no response rate anywhere in that. You do not need to know what percentage of 5,000 households will call. You need to know whether the handful who plausibly might is bigger than the number you need, and for a high-value job that comparison is usually settled before any rate is guessed. The cost of reaching 5,000 homes explains how a shared card gets to that price in the first place.

What the ROI calculator does, and what it does not

The calculator on this site does that sum forwards: it takes the households on a mailing, multiplies them by a response rate you choose from a short list, multiplies that by an average sale you set, and compares the result with the live price of the spot. The response rate is your guess, not a fact, and the calculator cannot make it one.

The honest way to use it is to pick the lowest rate offered and see whether the numbers still work. If they do at the bottom of the range, the mailing is a reasonable bet. It also counts one sale per customer, so a business with repeat visits should put a year’s worth in the sale field rather than a single ticket.

What about awareness, when the calls do not come?

Some of what a mailing does cannot be counted in six weeks, and the honest way to handle that is to name it without letting it become an excuse. A roofer may get few calls from a card in September and still be the name a household in Summerville remembers when a storm takes half the shingles off. That is real, and it is the reason the second and third cards to the same homes tend to do more than the first.

The test is whether awareness is being used to plan or to explain. If a first mailing produced nothing countable, ask first whether the offer gave anyone a reason to act now, because the offer is usually the problem before the medium is. If it produced a little and your category is one where the need arrives later than the card, mail again and keep counting. If you are explaining a third quiet mailing with awareness, stop. There is a longer answer in what a quiet first mailing tells you.

Attribution is imperfect, and that is fine

Attribution, working out which piece of marketing a customer came from, is never complete, and it does not need to be. A household in Mount Pleasant sees the card, searches your name, and clicks your Google listing; the listing gets the credit. Another keeps the card two months and then says “a neighbor” on the phone because they have forgotten. Both happen, and both mean you are undercounting the card rather than overcounting it.

The standard is good enough to decide, and the decision is whether to mail again. Two questions settle it. Did the countable response cover the cost of the spot? And did the part you could not count plausibly help? If the first is yes, mail again. If the first is no and the second is a genuine maybe, run one more with a better offer and the tally sheet in place. If the answer is no twice, put the money somewhere else and do not feel bad about it.

Whatever happens, keep the tally sheet from the first mailing. It is the only response rate that is actually yours, and the next card you plan should be planned from it.