Back to the on-screen lesson ·
A message that works names the reader's situation, promises an outcome, gives a checkable reason and a traced next step; two messages are compared by response rate through the same channel.
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.
You will build a message from the reader's situation, the promise, the proof and the next step, order its lines the way a reader takes them in, and compare two messages on one channel by response rate.
You have a segment and its situation, a promise, proof, and a channel that fits. A message is those four things in a few lines, in the order a reader needs them. This lesson writes it, and then shows how to tell whether a new version is really better than the old one.
| Term | What it means |
|---|---|
| Message | What the channel carries: a few lines a reader takes in quickly. |
| Situation | The line the reader recognizes as their own circumstance. |
| Promise | What will be different for the reader, stated plainly. |
| Reason to believe | Proof the reader could check: a record, a test, a name. |
| Next step | What to do now, with the channel's own code. |
| Response rate | Next steps taken over people reached, times a hundred. |
| Split test | Two messages through one channel, in equal numbers, at the same time. |
Most small-business messages are one of three things: a slogan ('Sparkle and shine!'), a list of services ('ovens, fridges, windows, bathrooms'), or a boast ('the best in the city'). Each fills space and none gives a reader in a hurry any reason to act.
A message that works is built from what the earlier lessons produced, in the order a reader takes it in. First, a line the reader recognizes as their own situation, 'Moving out this month?', because nobody reads further about something that is not for them. Then the promise: 'we clean to the property manager's inspection standard'. Then the reason to believe it, which is proof, not adjectives: '114 of our last 120 passed first time'. Last, what to do, with the channel's own code so the response can be traced: 'Book with code AGENT5'.
Four parts, a sentence or two each. Everything else, such as the full list of services or the history of the business, belongs on the page the next step leads to.
Another way: table
One message, taken apart.
| Part | Bright Home Cleaning's card |
|---|---|
| Situation | Moving out this month? |
| Promise | Cleaned to the inspection standard |
| Reason to believe | 114 of our last 120 passed first time |
| Next step | Book with code AGENT5 |
Another way: steps
Owners change their messages on a feeling. A better habit is a split test: put two versions through the same channel, in equal numbers, at the same time, and count responses. Northside Repairs put a slogan leaflet through 200 doors and a promise-and-record leaflet through another 200 on the same streets in the same week. The first brought 9 calls, a 4.5 percent response; the second 27, a 13.5 percent response, a gap of 9 percentage points.
Because the channel, the week and the numbers were the same, the gap belongs to the message. Change two things at once, such as a new message on a new channel, and you cannot tell which one mattered. A percentage-point gap is the difference between two rates; it is not the same as the difference between the two counts, and it is the number to compare when the two groups are the same size. With small groups, run the test again before trusting a small gap.
Start from the segment's own words. The last-time stories from the interviews give the situation line: 'we had the inspection on Friday' becomes 'Inspection this week?'.
State the promise as an outcome. What will be different for the reader, with the boundary from the promise lesson: 'ready Friday if dropped off by Thursday noon'.
Add one piece of proof. The kept rate, the test record, the named partner: something the reader could check. One strong figure is better than three weak ones.
Give one next step. One action, with the channel's code: call this number, scan this code, order by Thursday. Two steps split the reader's attention.
Split-test changes. When a new version is written, send it and the old one through the same channel, in equal numbers, at the same time.
Compare rates. Responses over people reached, times a hundred, for each version, and the gap in percentage points.
Check the answer. Read the message as a stranger in the segment: could you say in a sentence what is promised, why to believe it, and what to do? Check that each rate is below a hundred and that the two groups really were the same size and reached at the same time.
Opening with the situation is allowed, and needed, because readers decide in a second whether a message is for them. A question they answer yes to earns the next line; a slogan earns nothing.
Ending with one step is allowed because a message's job is to move the reader one step, not to close the whole sale. The page, call or visit the step leads to does the rest.
Comparing rates, not counts, is allowed because rates put both versions on the same footing. Twenty calls from 400 doors and fifteen from 200 look like a win for the first, but the rates are 5 and 7.5 percent. Only when the groups are equal do counts and rates agree.
Changing one thing at a time is what makes the comparison mean anything. If the message, the channel and the week all change, any difference could belong to any of them.
The four parts stay the same, but each channel gives them a different amount of room. A sign on a cycle path is read in two seconds from a moving bike: 'Flat? Fixed while you wait. 200 yards ahead.' A newsletter advert is read at a desk and can carry a full sentence of proof. An email subject line has a few words, so it carries the situation and nothing else; the promise and proof go in the first lines of the email.
Cut from the end of each part, never a whole part. A sign without proof can point to it ('ask to see our ready-on-time record'), and a subject line without a step leads to an email that has one. A message missing its situation, though, is read by nobody, and one missing its step is read and forgotten.
Every claim in a message is a promise the business must keep, and federal law agrees. The Federal Trade Commission requires advertising claims to be truthful and backed by evidence, and a figure like '114 of our last 120 passed first time' must come from real records. Keep the records that support each claim, update the figure when it changes, and never borrow a competitor's proof or round a 78 percent record up to 'nearly all'.
Honesty also sells. Customers have seen enough boasts to ignore them, and a precise, checkable figure stands out precisely because it could be wrong. A message that would embarrass the business if a customer checked it is not worth sending.
Small tests produce gaps that can come from chance. If 50 cards of each version bring 4 and 6 bookings, the gap is 4 percentage points, but one or two bookings either way would erase it. If 200 of each bring 16 and 40, a gap of 12 points, the second version is very likely better.
A simple rule for small businesses: act on a gap when the better version has at least twice the responses of the weaker and each has at least ten, or when the same gap shows up a second time. Otherwise, keep both running and count again.
Keep every message the business sends in one simple file: the date, the channel, the exact words, how many people it reached, and how many took the next step. After a year the file is a record of which situations, promises and proofs this business's customers respond to, and new messages can start from the winners rather than from a blank page.
The file also stops a common mistake: quietly going back to an old message that already lost a test. When someone suggests 'let's try the slogan again', the file shows what happened last time. And when a winning message starts to fade, with its rate falling month by month as the same readers see it again and again, the file shows that too, and it is time to test a fresh version against it.
A small heating and cooling company in Baltimore hung door hangers every spring that read 'Cool comfort, guaranteed! Installation, repair, maintenance, ductwork, filters.' Each run of 1,000 hangers brought about 8 calls. The owner, tired of paying for printing that did nothing, wrote a second version from what customers had told the office: most spring calls came the week of the first heat wave, from homes whose air conditioning had not been serviced.
The new hanger read: 'First hot week coming? Get your AC checked before it quits. 212 tune-ups last spring, 97 percent finished in one visit. Call this number by May 15 for a spring slot.' The phone number on it was used nowhere else.
The owner ran a split test, 1,000 of each version on matched streets in the same week. The old hanger brought 7 calls, 0.7 percent; the new one 31, 3.1 percent, a gap of 2.4 percentage points from the message alone. The next spring, every hanger carried the new message, and the company booked its spring tune-ups out three weeks ahead.
The Federal Trade Commission's guidance for businesses says that advertising must be truthful, must not mislead, and must be backed by evidence when it makes objective claims. A message built on a kept rate or a test record meets that standard naturally, because the proof is the claim's evidence.
A message needs only a catchy slogan. A slogan names no situation and gives no reason.
List everything we do, so nobody misses anything. A list asks the reader to do the work of finding themselves in it.
Proof is 'the best'. Proof is something the reader could check.
Change the message and the channel together to save time. Then nobody can tell which change worked.
Compare the counts. Compare rates, unless the groups are the same size.
Read the old advert.
$\text{Taste the difference! Salads, wraps, soups}$
A slogan and a list.
Write the situation.
$\text{Friday meeting?}$
The office manager answers yes.
Write the promise.
$\text{hot lunch for 8 to 20, by 12:15}$
An outcome with its boundary.
Add the proof.
$94 \text{ percent of the last } 200 \text{ on time}$
From the delivery log.
Add the next step.
$\text{order by 4 pm Thursday, code NEWS10}$
One action, traced.
Choose what changes.
$\text{the message only}$
Same streets, same week.
Send equal numbers.
$200 \text{ and } 200 \text{ doors}$
So counts and rates agree.
Count the calls.
$9 \text{ and } 27$
Slogan, then promise.
Work out the slogan's rate.
$9 \div 200 \times 100 = 4.5$
Calls over doors.
Work out the promise's rate.
$27 \div 200 \times 100 = 13.5$
The same division.
Take the gap.
$13.5 - 4.5 = 9$
Nine percentage points: keep the promise version.
Note the groups.
$400 \text{ and } 200 \text{ cards}$
Not the same size.
Note the bookings.
$20 \text{ and } 15$
The first looks better by count.
Work out the first rate.
$20 \div 400 \times 100 = 5$
Bookings over cards.
Work out the second rate.
$15 \div 200 \times 100 = 7.5$
The same division.
Take the gap.
$7.5 - 5 = 2.5$
The smaller group did better.
Check the counts are large enough.
$15 \text{ and } 20, \text{ both over ten}$
But not twice as many.
Decide what to do next.
$\text{run both again, equal numbers}$
A small gap needs a second look.
Write the situation the café owner recognizes.
$\text{Opening a café this spring?}$
A question they answer yes to.
Add the proof.
$\text{tested through 500 wash cycles}$
A record they could ask to see.
Add the next step.
Northside Repairs is writing a message for commuters with a broken bike. Which draft will do the most work?
Complete the worked solution: Northside Repairs put a slogan leaflet through $200$ doors and a promise-and-record leaflet through another $200$ on the same streets in the same week. The slogan leaflet brought $6$ calls and the promise leaflet $34$. Find each leaflet's response rate and the gap between them.
Find the slogan leaflet's rate.
$(\text{calls}) \div (\text{doors}) \times \text{a hundred} =$ p
Responses out of each hundred reached.
Find the promise leaflet's rate.
$(\text{calls}) \div (\text{doors}) \times \text{a hundred} =$ q
The same division for the second version.
Find the gap in percentage points.
$(\text{larger rate}) - (\text{smaller rate}) =$ g
The difference between the two rates.
Say why the gap belongs to the message.
$\text{same channel, streets, week and numbers}$
Only the message changed.
Maya's card for cafés reads as four lines. Match each line to the job it does.
| The reader's situation | The promised outcome | The reason to believe it | The next step | |
|---|---|---|---|---|
| Opening a café this spring? | ||||
| Cups that survive commercial dishwashers, delivered before you open | ||||
| Tested through $520$ wash cycles — ask to see the record | ||||
| Order a trial set with the code FIT20 |
Put the lines of Monica's board at the market entrance in the order a shopper should read them.
Number the steps in order (write the number in the box):
Northside Repairs put two versions of a leaflet through $200$ doors each. The slogan version brought $4$ calls; the version naming the Thursday-to-Friday promise and its record brought $32$. By how many percentage points was the second version's response rate higher?
Answer:
A pet groomer in Seattle emailed two versions of a reminder to $300$ customers each. The version with the subject 'Spring special!' brought $12$ bookings; the version with 'Is your dog's coat ready for summer? Book by Friday' brought $39$. If the better version had gone to all $1200$ customers on the list instead of the weaker one, how many more bookings would the groomer expect?
Answer:
A piano tuner is leaving cards at music schools for parents whose child has just started lessons. Which card will work best?
Lesson test: one question per skill, one attempt each, no hints. Your answers are checked when you submit.
Bright Home Cleaning gave the property managers $50$ cards of each of two messages. The card listing its services brought $7$ bookings; the card naming the inspection, the pass record and a booking code brought $13$. Fill in each card's response rate, as a percentage.
| Response rate, percent | |
|---|---|
| Services card | |
| Inspection card |
You can turn a slogan into a message that does four jobs and judge two messages with a fair split test. Tell someone why changing the channel and the message together teaches nothing. Next: where in the funnel people stop.
18. Your turn: Maya's shop-fitter card, step 3
$\text{order a trial set with code FIT20}$
One action, traced.