Measure marketing return

A campaign's return is the contribution from sales above a stated baseline over the campaign period, less its full cost; crediting it with every sale, or counting views, measures something else.

Paper packet. Every task here also exists on screen, where it is checked automatically; answers written on paper are not assessed by Nydus. When you are back at a device, enter your answers there.

1. What you will learn

You will measure a campaign's return from the extra units above a baseline, their contribution and the campaign's full cost, decide whether it paid for itself, and choose a baseline that does not credit the campaign with the season.

2. What you already have

You can compute contribution, acquisition cost and lifetime value. The last question of the course brings them to a single campaign: after it had run, was the business better off than it would have been without it?

3. Words for this lesson

4. Only what is above the baseline

$$\text{return} = (\text{weekly sales during} - \text{baseline}) \times \text{weeks} \times \text{contribution per unit} - \text{cost}$$

Neighborhood Kitchen usually sells about 140 lunches a week. For the 4 weeks its printed menu went round the office park, it sold 170 a week. Each lunch contributes 4 dollars; the printing and delivery cost 300 dollars. The campaign brought (170 − 140) × 4 = 120 extra lunches, contributing 120 × 4 = 480 dollars. Less the 300 it cost, it returned 180 dollars.

Now watch the most common mistake. Credit the campaign with every lunch sold in those weeks, 170 × 4 × 4 = 2,720 dollars of contribution, and it appears to have returned 2,420 dollars. But 140 lunches a week would have been sold anyway. The campaign is responsible only for what it added.

The same arithmetic can come out negative. Maya's paid posts lifted online orders from 10 to 16 a week for 6 weeks: 36 extra orders at 10 dollars of contribution is 360 dollars, against 450 dollars of cost. The campaign lost 90 dollars, however busy the shop felt.

Term	What it means
Baseline	What sales would have been without the campaign: the weeks before, or last year's season.
Incremental sales	Sales above the baseline during the campaign.
Incremental contribution	The contribution those extra sales made.
Full cost	Everything the campaign used: printing, fees, the owner's hours.
Return	Incremental contribution less the full cost; it can be negative.
Attention figures	Views, reach and likes: not money.

Another way: table

Neighborhood Kitchen's printed menu, measured two ways.

	Honest	Every sale credited
Lunches counted	120 extra	680 in total
Contribution	480	2,720
Less cost	300	300
Return	180	2,420

Another way: steps

Write the baseline down before the campaign starts.
Count weekly sales during the campaign.
Take the baseline off each week and multiply by the weeks.
Multiply the extra units by the contribution of one.
Take off the campaign's full cost.
Read the sign: above zero it paid, below zero it lost.

5. Choose the baseline carefully, and ignore attention

The whole answer depends on the baseline, so choose it with care. The weeks just before a campaign are a fair baseline in an ordinary month. They are not in December for a bakery, or in spring for a cleaner, when sales rise every year; then the fairer comparison is the same season last year, or a similar branch or street that the campaign did not reach. Write the baseline down before the campaign starts, as with any test, so that it cannot be chosen afterwards to flatter the result.

Views, reach and likes do not enter the calculation at all. They measure attention, and a campaign that was seen 9,000 times and sold nothing extra returned exactly minus its cost. Count the full cost, too: printing, fees, the owner's hours at their costing rate. A pattern in one business's records describes those customers, in that place, over that stretch of time. It is evidence for the next decision, not a promise about what any other market will do.

6. The method, step by step, and how to check it

Write the baseline first. Before the campaign: 'about 140 lunches a week, the average of the last six weeks', or 'last December's 90 loaves a week'. Note why it is fair.

Count sales during. Weekly, from the same records as the baseline, for the campaign's weeks only.

Find the extra units. Sales during less the baseline, each week, added over the weeks.

Turn them into contribution. Extra units times the contribution of one, not the price.

Take off the full cost. Printing, fees, samples, the owner's hours at the costing rate.

Read the result. Above zero, the campaign paid; below zero, it lost money, however busy the weeks felt.

Check the answer. The extra units must be fewer than the total sold in the campaign weeks. Contribution must be used, not revenue. The baseline must have been written down before the result was known. And if the return is large, ask whether something else changed in those weeks, such as a competitor closing or a holiday, that would have lifted sales anyway.

7. Why each step is allowed

Taking the baseline off is needed because the campaign can only be credited with sales it caused. The baseline is the best estimate of the sales that would have happened without it; anything above that is the campaign's work, as far as the records can tell.

Multiplying by contribution rather than price is needed because the extra sales still use up ingredients, supplies and time. Only their contribution is left to pay for the campaign.

Writing the baseline down first is what keeps the comparison honest. Chosen afterwards, a baseline can always be found that makes the campaign look good: the slowest week before, or a rainy month last year.

Allowing a negative answer is part of the method. A campaign that lost money is a useful finding: it says not to run that campaign again in that form.

8. Campaigns that keep paying

Some campaigns bring customers who come back, and their return does not end when the campaign does. Northside Repairs' free safety check brought 32 extra repairs in four weeks, but several of those riders became regulars. Counting only the campaign weeks understates what it returned.

The honest way to include this is to keep the two figures apart. Report the return in the campaign weeks, calculated as in this lesson. Then, separately, track the new customers it brought as a group, as in the retention lesson, and estimate their lifetime value, with the assumptions named. A campaign that loses a little in its own weeks but brings customers worth far more over the following year may be well worth repeating; the figures, kept apart, show it.

9. Keeping a campaign record

Write every campaign on one line of a simple record: the dates, what it was, the baseline and why it is fair, the weekly sales during, the contribution per unit, the full cost and the return. Add a note on anything unusual in those weeks.

Over a year or two the record shows which kinds of campaign pay for this business and which do not. Printed menus at the office park may pay every time; paid posts may lose a little every time. That is worth more than any general advice, because it is about these customers and this offer.

10. Finding a fair comparison

The weeks just before a campaign are the simplest baseline, but they are not always fair. Three other comparisons are often better, and each is within reach of a small business.

The same season last year. For a business with a strong season, such as a bakery in December or a lawn service in May, last year's same weeks carry the season in both figures. Adjust for any lasting change, such as a new product line, and say so beside the figure.

A place the campaign did not reach. If the flyers went to three streets and not to three similar streets nearby, the second group is a baseline for the same weeks. That is how the Albuquerque bookstore measured its mailer.

A code only the campaign carries. Sales with the campaign's code are a lower bound on what it brought, because some buyers will not use the code. Read together with a baseline, the code shows which extra sales came from the campaign and which from something else.

11. Return, payback and lifetime value together

The course's figures answer different questions, and a campaign is best judged with all three in view. The return in this lesson says whether the campaign paid for itself in its own weeks. Acquisition cost and payback, from the channels unit, say what each new customer cost and how soon they earn it back. Lifetime value says what a returning customer is worth over the whole relationship.

A campaign with a small negative return that brought many new customers with a high lifetime value may be a good investment. A campaign with a positive return that brought only one-time buyers may be the last of its kind. Keeping the figures apart, each with its assumptions named, lets the owner see which is which, and decide with the same honesty the whole course has asked for: evidence first, the decision second, and the assumptions written down where anyone in the business can check them.

12. In the world: an Albuquerque bookstore measures its holiday mailer

An independent bookstore in Albuquerque mailed a holiday gift guide to 4,000 nearby households every November, at a cost of about 2,400 dollars for printing and postage. December was always the best month, and the owners had always assumed the mailer was why.

One year they measured it properly. As a baseline they used the previous December, when a printing delay meant the guide went out late, to only half the households. They compared the books bought by customers in the neighborhoods that got the guide on time with those in the neighborhoods that got it late. The on-time neighborhoods bought about 260 more books over the season. At about 9 dollars of contribution a book, that was 2,340 dollars, against 2,400 dollars of cost: roughly break-even.

The owners did not drop the guide, because many of those buyers came back in the spring, but they cut the mailing to the neighborhoods with the most returning customers and added a coupon code to count it directly. The next year the smaller mailing cost 1,400 dollars and returned about 1,000 dollars above its cost in December alone.

13. In the world: comparison groups in marketing

Larger companies measure campaigns by holding out a comparison group, a set of customers or places that do not receive the campaign, and comparing their sales with those that do. The principle is the baseline in this lesson: the campaign is credited only with what it added.

14. Where this goes wrong

Credit the campaign with every sale in its weeks. Baseline sales would have happened anyway.

Use revenue, not contribution. Revenue from extra sales still pays for their own costs.

Choose the baseline after seeing the result. Then any campaign can be made to look good.

A negative return means the method is wrong. It means the campaign lost money.

15. Monica's radio mention

16. Northside Repairs' free safety check

17. Maya's paid posts

18. Your turn: Neighborhood Kitchen's printed menu

19. Guided practice

Before a spring-clean offer in the residents' newsletter, Bright Home Cleaning sold about $30$ cleans a week. During the $5$ weeks it ran, it sold $36$ a week. Each contributes $40$ dollars, and the campaign cost $500$ dollars. Did the campaign pay for itself?

20. Guided practice

Complete the worked solution: Neighborhood Kitchen usually sells about $101$ lunches a week. During the $2$ weeks of a printed-menu campaign it sold $125$ a week. Each lunch contributes $4$ dollars, and the campaign cost $152$ dollars. Find the extra lunches a week, the extra lunches in all, their contribution, and the campaign's return.

21. Guided practice

Maya's notes on a month of paid posts for her online shop. Mark every note that helps measure what the posts returned.

22. Practice

Before a free safety check for bikes brought in before summer, Northside Repairs sold about $12$ repairs a week. During the $4$ weeks it ran, it sold $20$ a week. Each contributes $25$ dollars, and the campaign cost $500$ dollars. Fill in the extra repairs the campaign brought and the contribution they made, in dollars.

23. Practice

Before a free safety check for bikes brought in before summer, Northside Repairs sold about $12$ repairs a week. During the $4$ weeks it ran, it sold $20$ a week. Each contributes $25$ dollars, and the campaign cost $500$ dollars. What did the campaign return after its cost, in dollars? (Negative if it lost money.)

24. Practice

A barbershop in Louisville always gets busier in the weeks before school starts. Last year, with no campaign, it did about $64$ haircuts a week in those weeks. This year it ran a $2$-week back-to-school campaign costing $229$ dollars and did $77$ haircuts a week. Each haircut contributes $18$ dollars. What did the campaign return after its cost, in dollars? (Negative if it lost money.)

25. Somewhere new

A bakery spent $197$ dollars on a campaign for its new sourdough in December, when it always sells more bread. How should it judge whether the campaign paid?

26. Lesson test

Lesson test: one question per skill, one attempt each, no hints. Your answers are checked when you submit.

27. Test question

Bright Home Cleaning usually books $36$ cleans a week. During a $5$-week window-cleaning add-on campaign it booked $39$ a week. Each clean contributes $10$ dollars and the campaign cost $367$ dollars. What did the campaign return after its cost, in dollars? (Negative if it lost money.)

28. What you can do now

You can say whether a campaign paid, and why crediting it with every sale gets the answer wrong. Tell someone why 9,000 views can mean a loss. This is the end of the course: the next course, Operations and people, is about delivering what these customers were promised.