Plan Meta ad tests around the offer, the creative and the buyer's stage

A Meta ad can earn a click from the right person and still lead nowhere. If the next page asks for a purchase or sales call before that person understands the offer, better creative may simply buy more expensive hesitation.

Work on the whole decision: who the offer serves, why they should care about it now, what the ad promises and what happens after the click. Then test the part you think is preventing a useful outcome.

Harry Long's Meta funnel walkthrough emphasizes an intermediate assessment for cold prospects. His review of well-known marketers' ads supplies several creative approaches to test. Both are useful starting material when their observations are kept separate from claims about proven performance.

Decide what the buyer can reasonably do next

Someone buying a familiar, inexpensive product may be ready to purchase after a clear ad. Someone considering a complex service may first need to understand the problem, the method, the supplier and the likely commitment.

That distinction changes the destination. Direct purchase can fit an understandable low-risk offer. A call can fit an urgent, well-defined need or a person already evaluating you. A guide, demonstration or assessment can fit somebody who needs help working out whether there is a problem worth solving.

Harry makes the case for the last group. The recipient gets an answer about their situation before being asked to join a sales process. The seller gets a more informed basis for a conversation.

Don't turn this into a rule that direct-response ads always fail. A useful intermediate step can also create unnecessary friction for someone already ready to act. Choose it because it helps the buyer make the next decision, not because every funnel needs more stages.

Write the decision down before commissioning creative. For example: “This campaign should help a small consultancy decide whether its lead qualification process needs attention.” That is a more usable brief than “generate cheap leads.” It tells you which people, questions, promises and results belong in the campaign.

Treat public ad research as a collection of hypotheses

The Meta Ad Library is a place to inspect publicly available advertising. It is not a competitor's private Ads Manager. Looking at a creative does not establish its spend, cost per qualified lead, revenue or profitability.

Harry describes reviewing ads associated with Alex Hormozi, Daniel Priestley, Gary Vaynerchuk, Grant Cardone and Neil Patel. The account below preserves his interpretation of their approaches. It does not independently establish that each pattern is used in every ad or that it causes better results.

The distinction matters when you borrow an idea. A famous marketer may have a large existing audience, a different offer, more cash to wait for a return, or a team capable of supporting a production volume that would overwhelm your business.

For an interesting ad, record who it appears to address, the opening, the specific promise, the evidence shown, the requested action and the destination. Add what you do not know. Then write a test in your own words with your own material. Copying somebody's creative assets or claims does not transfer their business conditions.

Give useful information inside the ad

Harry's Hormozi observation is a challenging opening followed quickly by usable information. The ad gives part of the answer before requesting a click.

A consultancy could test a short explanation of one qualification mistake and a check the viewer can perform. The destination can then help them examine the rest of the process. The opening needs a defensible point, not an exaggerated claim that every business has the same failure.

Harry also describes a large volume of creative variations in the library. That can suggest a testing habit; it doesn't tell you the budget, results or minimum number of ads you need. Produce enough distinct ideas to test sensibly within your own capacity.

Make the next step about the recipient

His Priestley observation is an invitation to a scorecard or assessment rather than an immediate sales page. The ad asks a useful question about the viewer's situation. The result page interprets the answers and offers an appropriate next step.

For a qualification service, that could be a short assessment of how enquiries are screened and handed to sales. The person should learn something useful even if they don't buy. A deliberately alarming score that everybody fails creates a misleading sales funnel, not a diagnosis.

Use a presentation style you can sustain

Harry describes Gary Vaynerchuk's advertising as more like raw or repurposed organic content, sometimes asking for a follow, subscription or community connection. The lesson is to test a credible, understandable presentation rather than assuming every ad needs a studio production.

Audience growth is also a different objective from near-term sales. It can require time and cash before producing a return. A small agency with immediate cash needs should not copy a broad awareness strategy and assume conversion will eventually take care of itself.

A natural-looking recording still needs a point, understandable sound and a truthful claim. Rough production is not a substitute for substance.

Address the relevant role and aspiration

In Harry's reading of Grant Cardone's ads, the opening speaks directly to the kind of person being addressed and the identity they want to build. That can translate into a straightforward creative choice: make the intended audience recognizable.

An ad for an agency delivery system could address founders who want to stop approving every routine task themselves. The body should explain what process or responsibility changes make that possible.

Avoid turning the callout into pressure about whether someone is “serious” or belongs to an exclusive group. A buyer can understand a relevant aspiration without being shamed into accepting the offer.

Lead with evidence when you have it

Harry's Patel observation is a researched, data-led opening. Specific evidence can be useful for a cold audience that doesn't yet know the speaker.

Use an actual finding, demonstration or permitted case. Explain the denominator, time period and conditions that change its meaning. If a lead count includes every form completion, do not describe it as qualified pipeline. If you cannot support a statistic, use a concrete demonstration instead.

The numbers mentioned as examples in the source review are not reusable facts. Specificity helps the reader inspect a claim; it doesn't make an unsupported claim true.

Build an honest assessment handoff

An assessment is worth testing when a few relevant questions can produce useful guidance and reveal whether the service fits. It is less suitable when the subject requires a professional diagnosis that a lightweight quiz cannot provide.

The landing page should explain who it helps, what they receive, roughly what effort is involved and how their answers and contact details will be used. Show a truthful example of the output where possible. A person shouldn't have to begin to discover that the result is only a booking invitation.

Ask questions that affect the result or a legitimate next step. For the consultancy example, these might concern how enquiries arrive, what qualifying information is collected, who reviews fit and what happens to unsuitable enquiries. Ask about budget or timing transparently when those answers serve a stated purpose.

Make the scoring explainable. Return useful recommendations for stronger as well as weaker results. A well-run business might need a small improvement, a resource or no service from you. Allow that outcome.

Then offer a next step that follows from the answers. A person with a clear need and suitable constraints can choose a conversation. Someone early in the process may prefer a guide or example. Someone outside the service's scope should not be pushed toward an irrelevant appointment.

The assessment design guide goes into the questions, results and follow-up in more detail. The ad campaign still needs to prove that this step attracts people the business can actually help.

Use the source results with their limits attached

Harry describes a parent-support assessment with about 1,100 starts and 1,000 completions, and a fitness assessment with 42 starts and 35 completions on its first day. In a separate lead-magnet video, he also discusses an FX trading assessment generating 252 leads.

Those are his reported client outcomes. They are not OutreachGenie results. The school and fitness examples are consumer services, and a completion does not establish that the respondent is qualified, attends a call or purchases.

The cost figures for the FX example differ across the source videos, so they should not become a planning benchmark. Nor should a completion rate among people who started a quiz be compared with a form conversion rate among all landing-page visitors. The denominator changes the question being answered.

Use one set of stages for your own measurement: page visits, assessment starts, completed responses, qualified opportunities, meetings held and customers. Preserve spend and time period alongside each count. A high completion rate is encouraging only if the earlier and later stages also work.

Make the first creative test answer one question

Harry suggests starting with a clear mockup of the assessment or its result on a phone or computer screen. That is practical because it shows what the person will receive. A tool such as Canva can help build the layout, provided the screenshot and promised output are accurate.

Use the copy to explain the viewer's problem or decision. “Find the weak handoff in your enquiry process” is more informative than “take our amazing quiz,” assuming the assessment can actually deliver that insight.

A short on-camera explanation is another option. Name who it is for, the question it helps answer and what the result contains. You don't need to explain the entire consultancy service before someone can decide whether the assessment is useful.

Keep the initial variants distinct enough to learn from. You might compare an example-result creative with a presenter explaining the decision. If both change the audience, offer and destination as well, you won't know which difference mattered.

Before launch, record the hypothesis. For example: “Showing a useful result should help suitable viewers understand why the assessment is worth completing.” Choose the outcome that tests it, such as qualified completed assessments per amount spent. Click-through rate can help diagnose attention, but it cannot settle lead quality.

Set the budget, review date and conditions for stopping. Those depend on the economics and risks of your offer. There is no source-backed universal minimum spend or predictable cost per lead to insert into the plan.

Give targeting and measurement enough care

Harry recommends avoiding an audience narrowed by excessive filters. The useful principle is to leave enough room to test inside the market you can serve.

Keep genuine constraints, including geography, eligibility, language and relevant buyer conditions. A broad audience for a local regulated service still needs boundaries. A niche B2B offer may face a small reachable market, while cheap clicks from a wider audience can conceal poor commercial fit.

Use actual downstream results to assess whether targeting is useful. Don't assume the platform has found your ideal customer because form completions become cheaper.

Verify measurement before paying for traffic. Check that the destination works, the promised result appears, events represent the stages you intend to measure and duplicate events are not inflating the count. Review current platform and assessment-tool documentation for the integration you actually use rather than following a stale interface walkthrough.

Collect and share data with the appropriate permissions. Don't send sensitive assessment responses, health information, school difficulties or financial details to advertising systems merely because a tool makes that possible. Track the necessary stage without exporting the respondent's private story.

Diagnose the stage that is failing

Give a properly instrumented test enough time and data to become interpretable. Harry suggests avoiding major conclusions from the first day and often reviewing after a week. That is an operating example, not a promise that learning ends in seven days or that cost will steadily fall.

A broken link, wrong audience, missing consent or runaway spend deserves immediate attention. Waiting for the algorithm is not a remedy for a setup error.

When the test is functioning, inspect the sequence:

Change the part your evidence points to. Replacing an ad is unlikely to solve an assessment whose first question doesn't work on mobile. Adding more traffic to a follow-up queue nobody handles only increases the waste.

Keep a record of changes and compare appropriate periods. A lower cost after a redesign may coincide with an audience change, a new offer or a different time window. Preserve that uncertainty rather than attributing every improvement to the last edit.

Use other channels where they help the same buyer

Harry's broader marketing training makes a useful point about B2B: buyers also use platforms outside LinkedIn. A relevant video, email or social post may help them understand the same question before they encounter an ad.

Repurpose the idea to suit that context. A short video can explain a single mistake. A longer guide can compare routes. A client newsletter can discuss a relevant operational change. They don't all need the same opening, asset or sales request.

Account for production work and permissions, including any follow-up or retargeting. Additional channels should make the decision easier for a suitable buyer, not make a refusal harder to escape.

At the next review, choose whether to continue, change or stop the test using qualified outcomes and cost. If the assessment produces inexpensive completions from people you cannot serve, tighten the offer and audience before celebrating the lead count. If it helps suitable prospects understand a real need and the economics work, you have a reason to invest in the next controlled experiment.