Hook rate and hold rate: what they are and how to use them in Meta Ads
Hook rate and hold rate for Meta video ads: how to calculate them, what values to watch, how to read them against CPA, and what native columns change.
71 posts
Paid media, CRO, measurement and growth. I write about what I run into on the job, without the agency filter.
Hook rate and hold rate for Meta video ads: how to calculate them, what values to watch, how to read them against CPA, and what native columns change.
Why Meta lead ads bring cheap leads that never buy, and the fix: qualifying questions, optimizing for conversion leads, and CRM feedback to Meta.
How to see visits from ChatGPT, Gemini, Perplexity and Copilot in GA4: a regex channel group, why some traffic is lost, and how to spot agents.
Education marketing in Latin America: the enrollment season, Google and Meta, how to filter inquiries, and why to measure enrollments rather than leads.
Google Ads for B2B SaaS startups: what counts as a conversion, feeding demos and deals back to Google, competitor keywords, and when to add LinkedIn.
Google Business Profile is a local business's cheapest source of customers. What to fill in, how to ask for reviews, and what costs you rankings in Maps.
How to advertise on Facebook and Instagram on a small budget: objective, targeting, how many ads, how much to spend, and why boosting posts is not enough.
Step by step: track WhatsApp button clicks with Google Tag Manager, mark them as a GA4 key event, use them in Google Ads, and connect chats to sales.
When a small business should hire a digital marketing agency, how agencies charge, the questions to ask before signing, and the warning signs.
How a real estate agency in Latin America should split ad spend across portals, Google and Meta, filter inquiries, and measure deals instead of leads.
Click-to-WhatsApp is the best-converting ad format in Latin America and the worst measured. How to set it up and see which chats become sales.
Most early-stage startups pay for ads too soon. What to have in place first, how to find the first channel, and when paid starts to make sense.
Advertising for clinics, dentists and practices: health restrictions on Google and Meta, which keywords to use, and how to track appointments instead of clicks.
The setting that costs local businesses the most money in Google Ads is on by default. Location, keywords, ad schedule and tracking, step by step.
There is no universal minimum Google Ads budget. How to calculate yours from cost per click and conversion rate, with a worked example.
For a small online seller in Latin America, the real question is who owns the customer. How to compare Mercado Libre Ads and Google Shopping.
A restaurant pays for marketing three times: app commission, ads inside the app, and its own ads. How to organize that money and measure what is incremental.
Order matters more than channel. What a local business or small company should do first with little money for ads, and what to leave for later.
First-touch, last-touch, linear, time-decay and data-driven, priced against the same conversion — so you can see what each one is biased toward.
Four platform changes landed in one week of September, and all four change what your numbers mean rather than what your business did.
The UNNEST pattern that unlocks event_params, SQL for sessions, funnels and first-touch vs last-touch, and how to keep the bill near zero.
Most of the work in Shopping is not in the ads platform — it is in the feed. That changes the staffing answer, and the usual split is not the one people expect.
Five layers can produce this symptom, and checking them out of order wastes weeks. The diagnostic sequence, starting with the two that cost nothing.
Seven checks that separate a specialist you can put in front of your client's account from one you will be apologising for.
Why ChatGPT Ads campaigns collect impressions and no conversions: impression-based billing, standard-only conversion events, and settings that lock at creation.
Frequency alone does not diagnose creative fatigue. The metrics that signal decay, how to tell it from audience exhaustion, and a refresh cadence that holds.
Most creative tests answer nothing. How to structure tests that isolate a variable, how long to run them, and the metrics that decide a winner.
Reach versus trust, and what each tier really costs. How to choose between macro and micro influencers by objective, and the maths that decides it.
Codes undercount, engagement means nothing, and last-click hides most of it. A measurement stack for creator marketing, from cheap signals to incrementality.
Video is not automatically better on paid social. How to decide by funnel stage, budget and production cost, and the test that settles it for your account.
Creator-style content is not automatically better. Where each format wins, what each really costs, and how to brief UGC that does not look like an ad.
Smart Bidding optimises for whatever you call a conversion. How to stop paying for junk leads: form hardening, offline imports and value-based signals.
The ROAS formula takes two seconds and gets misread constantly. How to work out break-even ROAS, why POAS and MER exist, and how to read all three.
Performance Max folds Search, Shopping, YouTube and Display into one campaign. When it genuinely performs, and the brand cannibalisation check to run.
The GA4 BigQuery export is not retroactive. What that means, how close the Data API gets you, and the schema differences that break joins.
How last-touch attribution assigns credit, a worked example, where it systematically misleads — B2B SaaS above all — and what to use instead.
Non-solicitation, account ownership, scope triggers, confidentiality and exit: the five clauses that decide whether a white-label PPC deal survives.
Hiring a Google Ads agency versus a freelance specialist: cost, seniority, coverage, and the situations where each one is the wrong call.
Reporting is where most white-label PPC arrangements quietly fail. The three models, what belongs in the report, and the handoff that keeps nuance.
Freelancer, white-label specialist, agency or offshore team: what each model of outsourcing Google Ads costs, and the six checks before you hand over.
The three pricing structures for white-label PPC, the rates each one lands at, and how to check your margin survives before you quote the client.
LatAm is not one market. What changes between Mexico, Argentina, Colombia and Chile when campaigns cross borders, and what breaks if you copy and paste.
For agencies: the real cost of hiring a paid search specialist versus using an external one, and the account volume where hiring starts to make sense.
White label PPC explained for agencies: what the services include, how an engagement runs step by step, pricing models, and who owns the client.
The GA4 BigQuery export is free and answers questions the GA4 interface cannot. What lands in the tables, the traps, and when it stops being optional.
Managing more than one market at a time completely changes how you have to think about growth strategy. A few concrete lessons.
Most businesses put almost all their effort into acquisition and almost none into what happens after. That's where most of the value gets lost.
Last-touch attribution gives all credit to the final click, overvaluing brand search and hiding what created the demand. What breaks, and what to use instead.
Acquisition cost alone tells you nothing. The relationship between LTV and CAC is what determines whether your growth is profitable.
Running the occasional A/B test isn't the same as having a CRO program. The difference is continuity and accumulated learning.
A simple way to decide how much to invest in Google, Meta, TikTok, and programmatic without guessing.
They're not competitors, they're different tools. When to prioritize Google Ads, when to prioritize Meta Ads, and when to use both.
Most marketing dashboards get abandoned within a few weeks. Why that happens and how to avoid it.
TikTok isn't just for consumer brands. How to think about the platform if your business is B2B or niche.
How programmatic buying actually works — the auction, the players, the formats — and the honest conditions under which it is worth adding to your mix.
A heatmap shows where people click and how far they scroll. The mistake is confusing that with why they do it.
Many A/B tests don't fail due to bad luck — they fail due to design mistakes that can be avoided from the start.
The principles of CRO are the same, but what to test and how to measure success change quite a bit between e-commerce and SaaS.
A growth plan that works isn't a list of ideas — it's a narrow set of prioritized bets with a clear way to measure whether they worked.
Choosing between PLG and sales-led is a structural decision that depends on how much value the product conveys on its own before anyone pays.
With small teams, the most important decision is what NOT to do. A simple prioritization framework avoids spreading effort across too many fronts.
A poorly chosen North Star Metric aligns the whole team around the wrong number. Here are the real criteria for choosing a good one.
GA4 and Firebase Analytics share one event engine. The one-line difference, when to use each, and why their numbers never match, with a setup checklist.
A dashboard with 40 metrics doesn't inform, it confuses. Its job is to answer three or four business questions fast, not show everything measurable.
With more ad blockers, Safari and Firefox limiting cookies, and iOS restricting tracking, measuring from the browser alone has become increasingly unreliable.
AI doesn't replace design or strategy judgment, but it completely changed how long it takes to get a site from idea to deploy.
Ad platforms already automated most of the tactical optimization work. The human value shifted toward strategy, creative, and business judgment.
A good prompt doesn't replace analysis, it accelerates it. These are the types of prompts that actually save time when reviewing campaign reports.
Hourly rate is the wrong criterion for choosing a consultant. This checklist helps evaluate what actually predicts a good outcome.
It's not a cost-per-hour decision, it's a decision about how much of that work you need available every day versus at specific moments.
The right interview questions filter far better than reviewing a resume full of recognizable platform names.