Case study templateInternal. Fill in after the first pilot.
Page 1 of 3: slots 1 to 3
The two-page case study,
and the interview that fills it in.
Pages 1 and 2 are the shape of the finished piece: five slots, with the rules for each. Page 3 is the interview that fills them in. Do the interview first.
Four rules, before you write a word
- Every percentage carries its denominator and its date: "87% of 64 enrolled workers on 14 March 2027", never "87% reach".
- Quotes are verbatim, from a named person with a title, approved in writing. Never draft one for somebody to sign off.
- Publish something that did not work. A case study with no negative reads as an advert and gets discounted.
- Get permission in writing for the words, the numbers, the titles and the logo, separately.
Slot 1 · Situation80 to 120 words. Their world before you turned up. No product words.
Headline and standfirst
Headline: one sentence, under 12 words, the outcome not the product. Standfirst: [Customer legal name], [sector], [region]. [N] sites, [N] enrolled workers, [N] languages, live since [month year].
The situation itself
Four things, in their words: how a change used to reach the site; who did that job; when they last had to evidence a briefing and what they handed over; what made them look. Name the trigger: an audit, a near miss, a new client contract, a crew who did not read English.
Slot 2 · What changed100 to 150 words, in the order it happened. What people did, not what the software does.
Week 0: what was set up, by whom, how long it took. Week 1: how people got on: where the join card went, who stood next to it, the activation number at day 14. Weeks 2 to 4: the first must-read that mattered, and the chase. Now: the habit. Who posts, how often, what gets exported.
Slot 3 · The three numbersThese are the case study; everything else is context. Denominator and date, or leave it empty.
Activation
Of the workers enrolled, the share who logged in within 14 days.
Numerator of
Measured on
Source: TeamTalk admin, users page
Confirmed reach
Of the workers one named must-read went to, the share who confirmed within 48 hours.
Which post
Published on
Source: the reach report for that post
Hours saved
The customer's own figure, worked out with them and agreed out loud.
Rounds a month
Minutes saved per round
Method: rounds × minutes ÷ 60. State it.
TeamTalk case study templatePage 2 of 3: slots 4 and 5, and getting it approved
Slot 4 · The quoteOne quote, 25 to 40 words. It must contain a specific thing that happened, not an opinion about the software.
“
Name Job title Approved by email on
Slot 5 · What we would tell another [housebuilder / caterer]Three bullets. Practical, transferable, and at least one of them a warning.
Write it as advice from one operator to another, in their voice where you can. The third bullet is the one that earns the trust: what they would do differently, or what still does not work and is still on paper. Name the sector in the heading, because a caterer will not read a housebuilder's case study.
Before you publishTick all eight. If one of them will not tick, the piece is not ready, however good the numbers are.
- Every percentage carries its denominator and the date it was measured, in the same sentence.
- The quote is verbatim, and the person has approved those exact words by email.
- Job title checked against their own email signature, not against what you remember.
- One thing that did not work is in the piece, in slot 5.
- Logo permission given separately and in writing, or the customer named in text only.
- They have seen the final PDF, not a draft, and replied.
- The figures are dated on the page, and you have diarised to re-check them at renewal.
- Nothing in it that you did not measure, or were not told by a named person.
Permission, in writingSend this within 48 hours of the interview, with the draft attached. Do not publish until it comes back.
Subject: your TeamTalk case study, for approval
[Name], thank you for the time. The draft is attached, along with the three figures exactly as we worked them out on the call: [activation]% of [N] enrolled workers on [date], [reach]% confirmed on "[post name]" within 48 hours on [date], and [hours] hours a year from [rounds] rounds a month at [minutes] minutes saved each.
Could you reply confirming three things separately: (1) the words and the quote are accurate and you are happy to be named as [job title]; (2) the three figures are right; (3) whether we may use your logo, or whether you would rather we named you in text only. If any figure is wrong, send me the right one and I will change it or take it out.
Nothing is published until you reply.
TeamTalk case study templatePage 3 of 3: the interview that fills in pages 1 and 2
Before the call
Book 30 minutes with the person who actually ran it, not their director. Ask to record it, and say why: "so I quote you correctly rather than from memory." Have their reach report and activation number open in front of you, because half the answers are numbers they will not have to hand. Send no questions in advance; you will get prepared answers instead of real ones.
On the call
Your job is to shut up. Ask the question, then wait. The useful sentence almost always arrives after a silence you were tempted to fill. When a number comes up, stop and get the denominator and the date before you move on; you will not get a second chance without looking sloppy.
A · Before (fills slot 1)
1
Before TeamTalk, when something changed on site, how did it reach everybody?Listen for: WhatsApp groups, a paper sheet, a supervisor's memory. Get the mechanism, not the intention.
2
Who actually did that job, and how long did it take them?You need a named role and a number of minutes. This is the input for the hours-saved figure later.
3
When did you last have to produce evidence that a briefing had happened, and what did you hand over?The best answers here are also the best quotes. Write the words down exactly.
4
What made you look for something different? Was there a particular week?This is the trigger. Keep asking until you have an event, not a sentiment.
B · Getting people on (fills the activation number)
5
Walk me through the first day. Where did the join card go, and who stood next to it?
6
How many people did you add, and how many were signed in by the end of the first fortnight?Write both numbers down. "Most of them" is not a case study.
7
Who was hardest to get on, and what worked in the end?Usually subcontractors or a night shift. This answer is what another site manager most wants to read.
C · The notice that mattered (fills the confirmed-reach number)
8
Which notice did you most want proof of? Tell me what happened with that one.
9
What was the confirmed reach on it, and how quickly?Get the post name, the date it went out and the denominator. Read them back to check.
10
What did you do about the people who had not confirmed?The chase is the story. Most buyers assume this is the bit that will not happen.
D · The time (fills the hours-saved number)
11
Take me through the old sign-off round: printing it, walking it round, chasing the ones who missed it, filing it. How many minutes, honestly?
12
And what does that look like now?
13
How many of those rounds do you run in a month?Do the arithmetic out loud with them: rounds × minutes saved ÷ 60. Get them to agree the annual figure on the call. A number the customer said is worth ten you calculated.
E · The quote, the warning and the referral (fills slots 4 and 5)
14
If a site manager at another [housebuilder / caterer] rang you and asked whether it was worth it, what would you say?This is your quote. Write the exact words, including the swearing you will have to take out.
15
What would you tell them to do differently from how you did it?
16
What has not worked, or what are you still doing on paper?Publish the answer. This is the sentence that makes the other three numbers believable.
17
Who else should I be talking to?Ask for two names. Referred leads convert far better than cold ones, and this is the moment they are most willing.