Proof of a real gap at every brand on your list
Obsession becomes a customer of each brand your client wants to win, watches what happens, and writes the facts back into your Clay table.
IllustrationThe agency, its client, the brands, the people and the numbers on this page are made up.
- Tidewren Swim
- Halvard & Moss
- Fennick Home
- Larkbound
- Its own inbox
- A UK mobile number
- Its own browser
- A control message and a second run
- Tidewren SwimNo texts after opt in
- Halvard & MossNo gap
- Fennick HomeNo welcome email
- LarkboundNo texts, no reply
Why installs aren’t enough
Install data tells you what a brand has, not what it does.
Everyone has the same signals
Store Leads and BuiltWith are open to every agency, so the brands you write to hear “noticed you use Klaviyo” every week.
Installed isn’t working
An install says the tool is there, not that it works. A mockup shows the fix, but nothing yet proves the problem.
Proof by hand doesn’t scale
Signing up, giving a number, waiting for replies and checking every inbox and phone, at 300 brands per client, is weeks of work for a small team.
Obsession gets you a dated fact about each brand, collected as a customer, about the exact problem your client’s product fixes.
Clay in, Clay out
It starts and ends in your Clay table.
No export, no new place to work. The brands go in from Clay and the facts come back to the same rows.
- Your list is already thereEach client’s TAM is a Clay table, built from your ecommerce data, Store Leads and BuiltWith.
- Facts come back as columnsUse them as a variable in your copy, a filter for who to write to, and a trigger when they change.
- Everything else still worksUpload a CSV, call the API or paste a list for one offs, like 25 brands before a pitch.
Step by step
From recipe to facts in Clay.
How the agency in this example runs it for one client, an SMS app, across 300 UK Shopify brands.
Set it up
Steps 1 to 5Create a watch
A watch is one list of companies, the checks to run on them, and how long to keep watching. Start one from a recipe, or by describing what you want to know.
In this exampleThe agency starts a new watch on SMS app prospects.
- Reviews app prospectsFrom the Prospect research recipe120 brandsWeekly
- Pitch: helpdesk prospectFrom the Speed to lead recipe25 brandsOnce
- Loyalty app prospectsDescribed in your own words210 brandsWeekly
Pick a recipe
A recipe sets up the checks for one job. Launch it as it is, change anything you like, or skip recipes and describe what you want to know in your own words.
In this exampleThe agency picks Prospect research. It proves a gap at companies you’re about to write to. Competitor tracking is for watching rivals over time.
Launch from a recipe, or describe what you want to know
- Competitor trackingRivals’ sign ups, emails, offers and launches, over time
- Prospect researchProof of a real gap at each company on your list
- Mystery shopperYour own or a client’s store, journey by journey
- Speed to leadHow fast and how well companies answer
- Price and promotion watchPrices, offers and terms as a customer sees them
- Ad trackingRivals’ ads, and where each click lands
- Trial teardownA rival’s whole onboarding, on one page
- Describe it yourself“Does each brand answer a sizing question within a day?”
Configure it
Choose the gap your client’s product closes and what counts as a gap. Each check has its own deadline, and fixed rules decide every verdict, so the same evidence always gets the same answer.
In this exampleEmail sign up, SMS opt in and one support question, watched for 48 hours, then every week.
Watching
The gap to prove
- Email sign up
- SMS opt in
- One support question
- WhatsApp opt in
- Live chat
- Basket, checkout and purchaseNeeds the brand’s OK
What counts as a gap
- No text within 24 hours of opting in
- No welcome email within 1 hour
- No support reply within 24 hours
- Recheck every miss from a fresh number before it counts
Watch each brand for
Afterwards
Shoppers based in
Alerts
Add the companies
Connect Clay, upload a CSV, call the API or paste a list. With Clay you choose a table and a view you’ve already filtered, point at the website column, and choose where each fact lands.
In this exampleThe agency connects Clay. It uses its “SMS popup” view, so no check is spent on a brand with no SMS at all, and turns on new rows so the watch grows with the table.
- Connect ClayConnected
- Upload a CSVOne column of websites
- APISend rows as you go
- Paste a listFor one offs
Bring companies in
- Table
- UK Shopify brands
- View
- SMS popup is yes
- Website column
- Domain
Send facts back
- GapNew column: Gap
- SeenNew column: Seen
- What happenedNew column: What happened
- ProofNew column: Proof
Rows are matched on the website column. Choose an existing column instead to fill that one.
Check and start
One summary of what will happen. Welcome emails can arrive within minutes, and each brand’s verdict lands 48 hours after its sign up.
In this exampleThe agency saves it as a recipe, “SMS app client”, ready to rerun for the next SMS client it signs.
- Recipe
- Prospect research
- Watching
- SMS app prospects
- Brands
- 300 now, plus new rows from the Clay view
- Checks
- Email sign up, SMS opt in, one support question
- Gaps
- No welcome in 1 hour · No text in 24 hours · No reply in 24 hours
- Watch
- 48 hours per brand, then every week
- Results
- Back to Clay as 4 columns, alerts in Slack
Obsession does the watching
Steps 6 to 8 · 48 hours in this exampleEach brand gets its own shopper
A real inbox, a real UK mobile number and its own browser, used for that one brand only. It signs up the way a customer would and saves a screenshot at every step. It always says it’s automated.
- Identity
- Declared as automated
- Inbox
- tidewren.4f2a@test.useobsession.com
- Mobile
- +44 7700 900418
- Browser
- Its own session, London
- Opened the store
- Found the sign up popup
- Signed up, ticked SMS consent
- Welcome email arrived, 38 s
- Control text arrived, 4 s
It waits, up to 48 hours
Every email, text and reply lands in that brand’s own inbox and phone. When a deadline passes with nothing, it tries again from a fresh number before calling it. Brands added later run on their own clock.
- Emails received
1,104
- Texts received
466
- Support replies
205of 296
- No text 24 hours after opting in
74of 296
- Halvard & Moss sent its second text
- Larkbound sent its third email
- Tidewren Swim: no text after 24 hours, second run started from a fresh number
- 12 new rows arrived from Clay and started their own watch
Results after 48 hours
A finding for every brand. A missed deadline only counts as a gap once a control message and a second run from a fresh number confirm it.
- Opted in for texts, none sent
61of 296
- Welcome email within an hour
231of 296
- No welcome email at all
19of 296
- Support took over 24 hours
84of 296
| Brand | Emails | Texts | Support | Finding |
|---|---|---|---|---|
| Tidewren Swim | 3 | 0 | 26 h | No texts after opt in |
| Halvard & Moss | 2 | 2 | 3 h | No gap |
| Fennick Home | 0 | 1 | 31 h | No welcome email |
| Larkbound | 4 | 0 | none | No texts, no reply |
296 of 300 finished. 4 had no working sign up form, which is logged as a finding too.
You get the proof
Steps 9 and 10Proof for every brand
The verdict, a timeline with every step’s time, the screenshots and the actual emails and texts. One link anyone can open, including the brand.
- Signed up with an email and a UK mobile
- Welcome email after 38 s
- Control text to the same number arrived in 4 s
- Asked support about sizing, declared as automated
- No text after 24 hours; second run from a fresh number
- Support replied after 26 hours
- 48 hours: 3 emails, 0 texts on either number
The facts land back in Clay
Four new columns on the same rows: the gap, the date it was seen, what happened in one plain sentence, and the proof link. Obsession reports what happened; your own Clay prompts decide what to say.
| Company | Contact | Gap | Seen | What happened | Proof |
|---|---|---|---|---|---|
| Tidewren Swim | Hannah PriceHead of CRM | No texts after opt in | 15 Sep | Opted in for texts on 15 Sep. 3 emails, 0 texts in the 48 hours since. | open |
| Halvard & Moss | Leo GrantFounder | No gap | 15 Sep | Welcome email in 2 min, 2 texts, support reply in 3 hours. | open |
| Fennick Home | Priya ShahHead of Retention | No welcome email | 15 Sep | Signed up on 15 Sep. 1 text, no emails in the 48 hours since. | open |
| Larkbound | Owen HaleEcommerce Director | No texts, no reply | 15 Sep | Opted in on 15 Sep. 4 emails, 0 texts, no reply to a support question. | open |
Obsession #sms_app_prospects
UK Shopify TAM: the first 300 brands are done. 61 opted in for texts and never got one, and 19 never sent a welcome email. New rows report as they finish. The facts and proof are in your Clay table.
Once the facts are in
What you do with the facts.
The opener
Our proof and your mockup, in one email.
The proof shows what’s happening at the brand today. Your mockup shows what it could look like with your client’s product.
From Sam, SMS app to Hannah Price, Tidewren Swim · Subject Tidewren’s texts
Hi Hannah,
We opted in for Tidewren’s texts on 15 September. Three emails arrived in the next two days, but not one text.
I’ve mocked up what your first three could look like, using your own welcome offer. Worth a look?
Sam
Proof: Tidewren, 15 to 17 SepMockup: Tidewren texts
Written by your own Clay prompt, from the “What happened” column.
Your mockup
TidewrenText message
Day 0
Welcome to Tidewren! Here’s 10% off your first suit with code TIDE10.
Day 2
The Ardley is back in your size. Want us to hold one for you?
Day 5
Last day for your 10%. Reply with your usual size and we’ll suggest a fit.
A sharper list
Write only where there’s a gap
Filter the Clay table on the Gap column and leave the rest out of the sequence.
Example61 of 300 brands get the SMS email this month.
Triggers
Changes become reasons to write
Weekly checks tell you when to write, and when to stop.
ExampleTidewren sent its first text: pause the sequence.
Something to sell
A monthly add on per client
Brand evidence for each SaaS client: brands checked, gaps open, gaps fixed, a proof link for each.
ExampleClient A, October: 73 open gaps, 12 fixed.
Winning clients
Open the pitch with proof
Paste 25 brands from a prospect’s TAM before the first call.
Example“Nine of 25 brands you’d sell to took a text opt in and never sent one.”
About this page: it illustrates how the product works. The agency, its client, the brands, the people, the phone number and every figure are made up, and no real company’s results are shown. Basket, checkout and purchase checks run only with the brand’s OK. Clay, Store Leads, BuiltWith and Klaviyo are trademarks of their owners, named here to describe where data comes from and goes.
Try it on one client.
Join the waitlist, and we’ll run prospect research on 20 brands from one client’s list, with the facts back in your Clay table.