The Summit resource library

Plan your next campaign with confidence.

Practical tools for better marketing decisions.

Explore campaign planning, creative development, and performance strategy.

Plan the test. Set a clear objective and a useful measure of success.
Shape the creative. Find a stronger opening and build the idea around it.
Practice the decision. Explore campaign scenarios before committing a budget.

Explore the exercises Open the live slides

Your next move starts here

Get the campaign toolkit

Build a focused campaign plan, develop stronger creative, and make better decisions about your ad budget.

Put your plan into practice

Make the next
decision count.

Work through a real marketing question before you put a budget behind it.

Fix the funnel Three businesses are stuck. Find the actual blocker. 2 min each
Choose the objective Match the business question to a signal you can see. 2 min
Find the hook One offer, three openings. Pick the one worth testing first. 2 min

Workshop exercises

Open the practice lab.

Complete the short resource form to explore these exercises and take the workshop materials with you.

Open the workshop

01 2 min each

Fix the funnel

Start with what you can observe. Then choose a test that answers the business question. These three fictional businesses give you room to practise.

Fictional workshop scenario Adapted classroom exercise. All names, businesses and figures are fictional; amounts are planning examples, not forecasts.

Priya Raman

Built an app that does the tax and budgeting maths for freelancers

I just need to find the right people to target.

Sells: Free download, then $8 a month after a 14-day trial
Where she is: 400 downloads, all from friends and one Reddit thread
Wants: 500 downloads a month
Her plan: $300 for the month, launching on Facebook, Instagram, TikTok, Reddit and LinkedIn at the same time
How she measures: Checks the App Store dashboard every morning and counts installs
For this exercise: No app event integration has been verified. Priya can run a click test while she investigates it.

1 / 3 Name the constraint

What is actually blocking her?

Her own answer is targeting. Read the facts again before you agree.
- Her audience needs a tighter interest stack.
  
  Not this one
  
  Audience choices matter, but narrowing them does not create a missing install signal. Verify measurement before judging an acquisition campaign.
- She has no verified app-install signal for the ad platform.
  
  Strongest choice
  
  An App Store total tells Priya how many installs occurred. It does not, by itself, verify ad attribution or provide the platform with an install event to optimise against. She needs to check her app measurement setup.
- The creative needs replacing before anything else.
  
  Not this one
  
  There is not enough evidence to blame the creative. A click test can compare hooks, but an install campaign needs a verified install signal first.
From the public workshop · Slides 32 and 40
2 / 3 Choose the next test

She has $300 and one month. What is the next test?
- Split the $300 across five platforms this month.
  
  Not this one
  
  That leaves $60 per platform for the month. Five different auctions and creative contexts would make a small test harder to interpret. Check the chosen platform’s minimum budget before planning any paid test.
- Launch an install campaign before checking app measurement.
  
  Not this one
  
  The test cannot answer an install question until the event setup is verified. Finishing measurement may be the most useful work this month.
- If budget minimums allow, test three hooks on one platform using clicks.
  
  Strongest choice
  
  This is a creative test Priya can label honestly. It can compare response to three ideas while she checks app measurement. If the budget cannot support a useful test, hold it and use interviews or organic creative feedback first.
From the public workshop · Slides 26 and 31
3 / 3 Choose the metric

For the click test she can run now, what does she judge?
- Total installs in the App Store dashboard.
  
  Not this one
  
  This total mixes sources and does not attribute an install to a tested ad. Keep it as business context, not proof of this traffic test.
- Cost per install, even though the install signal is unverified.
  
  Not this one
  
  That would imply measurement she has not established. Once install attribution and paid conversion are verified, she can evaluate acquisition economics.
- Clicks and cost per click by hook, with downstream installs explicitly unproven.
  
  Strongest choice
  
  For this traffic test, record clicks, click rate and cost per click alongside each hook. Decide the review date and spending limit first. A click result is useful creative evidence; it is not install or paying-customer evidence.
From the public workshop · Slides 31, 45 and 47

The immediate lesson: separate a creative click test from an install campaign. Verify measurement and customer economics before treating clicks as acquisition.

Fictional workshop scenario Adapted classroom exercise. All names, businesses and figures are fictional; amounts are planning examples, not forecasts.

Dani Ruiz

Runs a lash and brow studio on her own in East Austin

The videos are working. I just need better targeting.

Sells: $180 a set, and a regular comes back every three weeks
Wants: Twelve new clients a month. She has room for twenty
Does now: Four Reels a week, and hits Boost on the best one for $75 a time, about $400 a month
Last month: 20,000 views, ninety comments, three new clients, and all three came from a friend telling them about her
How you book her: Instagram DM. No website

1 / 3 Name the constraint

Twenty thousand views and three clients. Why?
- She is buying engagement without a measured path to a qualified inquiry.
  
  Strongest choice
  
  Views and comments show response to the video. They do not establish booking intent. Her DM route exists, but she needs a clear next action and a way to connect inquiries with bookings.
- The videos are definitely the problem.
  
  Not this one
  
  The videos attracted attention. We still need to test whether the offer and call to action attract local people who want an appointment. Views alone cannot prove either outcome.
- A smaller local radius will solve the whole funnel.
  
  Not this one
  
  Dani should restrict delivery to an area she can actually serve. That geographic constraint is useful, but it does not replace an inquiry objective, clear offer or booking follow-up.
From the public workshop · Slides 12 and 31
2 / 3 Choose the next test

She has about $400 a month and no website. Where does that money go?
- Keep boosting for engagement and judge it by comments.
  
  Not this one
  
  That repeats the current test. If her goal is appointments, she needs to evaluate qualified inquiries and booked clients, not only engagement.
- Test a Leads campaign with an instant form and a clear booking follow-up.
  
  Strongest choice
  
  An in-platform form can collect inquiries without a website pixel. Ask only what is needed to follow up, include a relevant qualifying question, and meet the platform’s lead-ad terms and privacy requirements before launch.
- Build a booking website before exploring any other intake route.
  
  Defensible, not the strongest
  
  A booking site may improve the full journey. It is not required for every lead test: an instant form or a measured messaging flow can be considered while that work is scoped.
From the public workshop · Slides 05, 30 and 36
3 / 3 Choose the metric

She wants twelve new clients a month. What is the number that tells her?

A set is $180 and a regular returns about every three weeks.
- Views and comments on the Reels.
  
  Not this one
  
  These show attention, not whether an inquiry became a client. Keep them as supporting creative measures.
- Cost per qualified inquiry and per booked new client.
  
  Strongest choice
  
  Track form submissions through qualification and booking. Her $400 example budget would need a cost of about $33 per booked client to produce twelve. That is a planning requirement, not a forecast. Check whether margins, capacity and actual conversion rates support it.
- Follower growth on the account.
  
  Not this one
  
  Followers may matter for future reach, but they do not answer the twelve-client question. Use the measure connected to the next business action.
From the public workshop · Slides 31 and 47, workbook page 1

Dani has attention. Her next test connects a clear offer to a qualified inquiry, then follows it through to a booked client.

Fictional workshop scenario Adapted classroom exercise. All names, businesses and figures are fictional; amounts are planning examples, not forecasts.

Marcus Obi

Bottles cold brew coffee and sells it at two farmers markets

That traffic was junk. I need to find real buyers, not tyre kickers.

Sells: $34 for a four pack, through a Shopify store
Wants: A hundred online orders a month. He gets eleven, and every one met him at a market
What he did: One campaign in June. Picked Traffic, sent everyone to his home page, $30 a day for two weeks
What happened: Nine hundred visitors, four orders. He switched it off
On his site: He has never installed anything. “It is Shopify, so I assume it tracks”

1 / 3 Name the constraint

Nine hundred visitors and four orders. What should he check first?
- The audience must be low quality.
  
  Not this one
  
  That is one possible explanation, but the facts cannot isolate it. Review the optimisation goal, event setup and landing page before concluding that the audience is wrong.
- He bought traffic, while purchase measurement and the landing journey are unverified.
  
  Strongest choice
  
  A Traffic objective seeks visits or clicks. Marcus wants orders. First verify purchase events and the product-to-checkout experience; then choose an optimisation goal aligned with the next test.
- The budget alone proves the campaign could never work.
  
  Not this one
  
  The budget constrains how much can be learned, but these figures do not prove it caused the outcome. Separate budget limits from measurement, offer and page problems.
From the public workshop · Slides 30, 32 and 40
2 / 3 Choose the next test

What does he run next?
- Switch platforms because coffee is popular there.
  
  Not this one
  
  Popularity does not show where this offer will work. Choose a channel for a reason you can test and fix the measurement gap before comparing acquisition results.
- Verify purchase events and the checkout, then plan a purchase-focused test.
  
  Strongest choice
  
  Use the supported store integration and verify events, consent settings and attribution before spending. A focused product page gives the offer a clearer destination. Setup time and available settings depend on the platform and store.
- Keep testing clicks, but use a specific product page.
  
  Defensible, not the strongest
  
  This can answer a limited landing-page or creative question. Label it that way and verify downstream measurement; cheap clicks alone do not establish profitable sales.
From the public workshop · Slides 31 and 32
3 / 3 Choose the metric

What measure would answer the purchase question?
- Sessions and click-through rate alone.
  
  Not this one
  
  These help diagnose the ad-to-page step. Add measured purchases and acquisition cost to evaluate his sales goal.
- Measured purchase cost against a margin-based limit and a written review plan.
  
  Strongest choice
  
  Include product margin, fulfillment and repeat purchase assumptions when setting the limit. Monitor for broken tracking or checkout immediately; review performance at a planned point with enough data to interpret.
- Impressions and reach alone.
  
  Not this one
  
  Awareness can be a separate goal, but it does not answer this purchase test. Match the report to the question Marcus is trying to answer.
From the public workshop · Slides 20, 46 and 47

A clearer product destination and verified purchase measurement make the next test easier to interpret. They do not guarantee more orders.

02 2 min

Choose the objective

Three adapted classroom puzzles. Choose what to optimise for now, then name what that result still cannot prove.

Fictional workshop scenario Adapted classroom exercise. All names, businesses and figures are fictional; amounts are planning examples, not forecasts.

1 / 1 Match the objective to the signal

Which objective fits this question?
- Video views, with a defined view measure.
  
  Strongest choice
  
  That matches the immediate attention question. Define what counts as a view and compare the creative on the same basis. A view does not prove awareness lift or a booking.
- Purchases, even though no purchase event is available.
  
  Not this one
  
  A purchase objective cannot answer this setup without the required conversion configuration. First define and verify the signal you intend to use.
- Reach, if the real question is how many people saw it.
  
  Defensible, not the strongest
  
  Reach is a useful awareness measure when exposure is the goal. It answers a different question from how long people watch a hook.
Attention test

Business question

Does the opening earn attention?

Objective

Video views

Signal

A defined platform view measure

Data source

Platform reporting

Next check

Does attention lead to qualified interest?

Platform labels and available view measures vary. Keep the comparison consistent and treat downstream business outcomes as a separate question.

From the public workshop · Public deck: campaign objectives and upper-funnel examples

Optimise for the question you are testing, then keep the limits of that signal visible.

Fictional workshop scenario Adapted classroom exercise. All names, businesses and figures are fictional; amounts are planning examples, not forecasts.

1 / 1 Match the objective to the signal

What can this test honestly optimise for?
- Traffic, optimised for clicks.
  
  Strongest choice
  
  Clicks are the verified signal in this exercise. Use the result to compare response to the hooks. It does not tell you which ad caused an install or acquired a paying customer.
- App installs, with no verified app measurement.
  
  Not this one
  
  This skips the known blocker. Check the app integration and attribution before making installs the optimisation event.
- Video views, to compare attention instead.
  
  Defensible, not the strongest
  
  That is a valid different question. It can test attention to the opening, but it does not answer the click-response question Priya chose here.
A bounded creative test

Business question

Which hook earns a click?

Objective

Traffic

Signal

Clicks

Data source

Platform reporting

Not established

Install or paying-customer attribution

A cheap click is not a cheap customer. Decide the budget, review point and comparison rules before starting.

From the public workshop · Public workbook: Priya; public deck: measurement and Traffic

A narrower question can still produce useful evidence. Keep the claim as narrow as the measurement.

Fictional workshop scenario Adapted classroom exercise. All names, businesses and figures are fictional; amounts are planning examples, not forecasts.

1 / 1 Match the objective to the signal

What is the clearest next event to optimise for?
- Leads, using instant-form submissions.
  
  Strongest choice
  
  The form gives the platform a measurable inquiry event without a website pixel. Ask a relevant qualifying question and connect submissions to a real follow-up process. Complete the platform’s terms and privacy setup before launch.
- Engagement, with comments as the success measure.
  
  Not this one
  
  That repeats the attention question. Comments do not establish which local people want an appointment.
- Traffic to a future booking page that is not ready.
  
  Not this one
  
  The destination must exist and work before visitors are sent there. An instant form is one practical intake route while the site is being developed.
Inquiry test

Business question

Who wants to discuss an appointment?

Objective

Leads

Signal

Instant-form submission

Data source

Platform form reporting

Business check

Qualified inquiries and booked clients

A form submission is not automatically a good lead. Review relevance and the booking rate, not only the cost per form.

From the public workshop · Public workbook: Dani; public deck: instant-form lead generation

The platform can optimise for a submission. Dani still needs to qualify it, follow up and measure the booking.

03 2 min

Find the hook

Compare three openings for the same fictional offer. Choose a credible angle worth testing, then write a version of your own.

Fictional workshop scenario Adapted classroom exercise. All names, businesses and figures are fictional; amounts are planning examples, not forecasts.

1 / 1 Pick the strongest test

Which of these is the strongest first test?

Use the stated audience, a specific problem and a claim the offer can support.
- “Meet Ledgerly, the all-in-one financial companion for independent professionals.”
  
  Not this one
  
  This opening describes the fictional product before giving a problem-aware freelancer a reason to care. It is a testable angle, but it is less specific to the stated audience than the problem opening.
- “Quarterly taxes again. Did you set anything aside?”
  
  Strongest choice
  
  A concrete situation and a direct question make this a useful first candidate for the stated audience. It does not guarantee attention or conversion. Compare it with a different credible angle and check actual results.
- “Never owe another surprise tax bill, guaranteed.”
  
  Not this one
  
  The fictional offer does not support this guarantee. Remove the absolute claim and test a benefit the product can demonstrate. Strong wording still needs evidence.
A useful creative comparison

Audience

A problem-aware freelancer

Specificity

A concrete quarterly-tax situation

Credibility

No unsupported tax-outcome guarantee

Variation

A different opening idea for the same offer

Evidence

Actual test results, not a predicted winner

For the next variant, keep the offer and audience stable and change the angle. Use the test to learn which idea deserves more work.

From the public workshop · Hook Sheet pages 1 and 2, slides 28 and 29

This is a candidate to test, not a guaranteed winner. The evidence arrives when real people respond.

Now write one

Your opening line, for the same offer

Five writing prompts inspired by the Hook Sheet, checked in your browser as you type. These simple wording checks cannot verify a claim or predict performance. Read the suggestions, then test the idea. Your draft stays on this page and disappears on reload.

What the checklist looks at