Back to Blog
Blog

You’re paying your marketing agency. But how do you know it’s working? We’re spilling secrets here.

October 15, 2026

You opened the email and you see that your agency has sent you last month’s reports. Colourful slides, big numbers, graphs going up and to the right. 

10,000 impressions. 4,200 reach. Follower count up 12%.

And did you catch yourself thinking ‘so what’?

The clarity you really want from your agency reports is not dazzling numbers that you aren’t sure means anything. It is the answer to a simpler, basic question: ‘What did marketing do for our business this month?

Not what they published. Not what they optimised. Not what they’re planning. They’re all interesting, sure, but what did it do for you — for your pipeline, your visibility, and your revenue?

Let’s discuss how to measure the success of your marketing efforts so you can understand how well your agency is doing its job. Or, as it is bound to happen in some cases, find out if you’ve been oversold by your agency.  

Ad campaigns are probably the most easiest functions to monitor. When ads are underperforming, the leads dry up and it’s pretty obvious from the numbers that it’s going well. But what about everything else? 

What about the SEO retainer you’ve been paying for eight months? The website your agency built six months ago? The content they’ve been publishing? The GEO strategy they pitched you on?

These are slower, quieter investments — but if you don’t know what good looks like, you can’t tell the difference between an agency that’s working and an agency that’s waiting out the clock. I don’t want to sound like agencies are dishonest and are always trying to trick you with the reports.  This is by no means an attack on agencies. We are one, and most agencies I know work hard and care about results. But our industry has its share of players  who promise the moon in the pitch and then hid behind “it takes time”. I’ve seen it from the inside, and I’d rather you knew what to look for. 

This is a founder’s guide to measuring the success of SEO, GEO, website design, content marketing, product marketing, and ad campaigns — without a marketing degree. 

In this article, we’ll go over: 

  • How to spot a vanity metric?
  • SEO takes time. How much time is too much?
  • Can anyone guarantee you’ll show up in ChatGPT? (GEO and AEO, explained honestly)
  • When anyone can write with AI, how do you tell if your content marketing is any good?
  • Your new website looks great. But is it bringing in business?
  • What does product marketing actually do, and how do you prove it worked?
  • What should you track in Google Ads, and what should you ignore?

First, learn to spot a vanity metric

Vanity metrics feel significant but say almost nothing about whether your business is growing. Impressions, reach, follower count, likes, ‘brand awareness’, and page views with no context, are all examples of what one might call vanity metrics. 

It’s not that these numbers don’t mean anything or that anyone can make these numbers grow without trying – that’s not true at all. They are early signals, especially for a new brand, but they are more like a warm-up act and less the final results. 

Impressions don’t pay salaries and if your agency reports are highlighting just that, you many have a problem. Many agencies are set up to manage accounts, not businesses, so they optimise for keeping you happy enough to renew with impressive-looking results. A conversation about leads, conversion rates and revenue is harder than a happy update saying you got 10% growth in your views, but guess which one your business needs? 

The resulting irony is absolute: the very agencies most obsessed with broadcasting platform metrics are the ones losing clients the fastest. According to this Forbes article, PPC-focused agencies that anchor their reporting on vanity metrics suffer a staggering 49% annual churn rate. In stark contrast, full-service agencies that tie their worth directly to business outcomes cut that turnover completely in half, sitting at just 25%. The agencies trying to protect their business with superficially impressive data end up losing the businesses. They are basically firing themselves. 

Whatever services you are paying for, ask for and review the cost per qualified lead (not cost per click), lead-to-close rate (if it’s low, marketing is attracting the wrong people), returning visitors and engagement rates (signals genuine interests), share of voice in your category (do you come up or when buyers ask AI about your space) and the actual revenue that was influenced by marketing efforts. 

Try this with the last report your agency sent. Cross out every metric that doesn’t connect to pipeline or revenue. Look at what’s left. I hope it’s better than a wry ‘not much’! 

Yes, SEO takes time. How much time is too much?

Any honest agency will tell you that meaningful organic results take anywhere from three to six months in most categories, longer in competitive ones. That’s why SEO is usually sold as a retainer, not a one-off project.

But SEO retainers shouldn’t be a blank cheque.

What your agency should be reporting, and when:

Within the first 60 days, you should see evidence of foundational work — a technical audit completed, on-page fixes implemented, a keyword strategy you’ve actually been shown and agreed to. If two months in, you haven’t seen a document, a plan, or a single explained decision, that’s a problem.

By month three, you should see early movement — not necessarily rankings, but signals. Pages being indexed. Keyword positions appearing, even if they’re on page three. Organic impressions increasing in Google Search Console.

By month six, you should be having a real conversation about which keywords you’re ranking for, which pages are driving traffic, and what the plan is for the ones that aren’t.

Not all traffic is equal

Fifty visitors from long-tail, bottom-of-funnel searches can be worth more than five thousand from one broad, generic keyword. If your report says ”Organic traffic is up 40%”, it looks good but means nothing until you know which searches brought those people in. Keywords sit at different stages of a buyer’s journey:

  • Top of the funnel: broad, informational searches like “what is accounting software.” Huge search volume, lots of curious people, very few ready to buy.
  • Middle of the funnel: comparison searches like “best accounting software for small businesses.” Less volume, but these people are actively evaluating options.
  • Bottom of the funnel: searches with clear buying intent, like “GST accounting software for restaurants pricing” or “[your brand] vs [competitor].” Low volume, high value. These people are close to a decision.

Many bottom-of-funnel searches are long-tail keywords: longer, specific phrases that few people search for, but the ones who do know exactly what they want. 

This is why a traffic graph alone tells you very little. An agency can grow traffic fast by ranking for easy, high-volume informational terms that will never bring you a customer. Good SEO deliberately balances all three stages: top-of-funnel content to build visibility, and middle and bottom-of-funnel pages to bring in buyers.

How to check it yourself

Ask for access to Google Search Console and Google Analytics, or ask your agency to show you both on a call. In Search Console, the queries report tells you which search terms bring people to your site. In Analytics, look at which landing pages lead to enquiries, sign-ups or purchases. Put the two together and you’ll see whether your traffic is made up of buyers or browsers.

The question to ask: “Which keywords are we targeting at each stage of the funnel, where do we rank for them today compared with three months ago, and how many enquiries came from organic search this month?”

If your agency can’t answer that question clearly, in plain language, without a forty-slide deck — you have a problem.

Can anyone guarantee you’ll show up in ChatGPT? (GEO and AEO, explained honestly)

GEO and AEO are two terms you’ll hear a lot right now:

  • GEO (Generative Engine Optimisation) is about getting your brand mentioned and cited when people ask AI tools like ChatGPT, Perplexity, Gemini or Google’s AI Overviews about your category.
  • AEO (Answer Engine Optimisation) is about structuring your content so that any engine, AI or traditional search, can lift a clear, direct answer from it. Think featured snippets, “People also ask” boxes and voice assistants, as well as AI answers.

They overlap heavily, and a good strategy covers both. But let me be straight with you: this is genuinely new territory. Nobody, including us, has it fully figured out. Any agency that tells you otherwise is overselling.

Why one check proves nothing

Have you noticed that AI answer engines don’t give the same answer every time? Ask the same tool the same question twice and you may get two different responses, with different brands mentioned. Results also shift with how the question is worded, the user’s location, their past conversations and updates to the model itself. 

So if you typed your category into ChatGPT once, didn’t see your name, and concluded the GEO work has failed, that isn’t a fair test. And no one can guarantee that your brand will appear in AI answers. If someone promises that, walk away.

There are now plenty of tools that track AI visibility, and some established SEO platforms have added AI tracking features too. They’re useful, but treat them as experimental. Most work by running sample prompts and estimating how often you appear, so the numbers are directional, not exact.

What you can realistically track

• Mention rate over time. Agree on a fixed set of 15 to 30 questions your customers actually ask. Run them across several AI tools, several times each, every month. Track how often you’re mentioned compared with competitors. The trend matters more than any single result.

• AI referral traffic. Visits from ChatGPT, Perplexity and similar tools show up as referrals in Google Analytics. Small numbers today, but worth watching.

• What your leads tell you. Add “How did you hear about us?” to your enquiry form. More and more buyers are answering “ChatGPT.”

• Content built to be cited. Clear structure, direct answers near the top of the page, real data and original insight. Ask to see specific pages that were created or rewritten for this.

• Your wider presence. AI tools lean on what the rest of the internet says about you: credible publications, reviews, directories, podcasts, LinkedIn. Ask what your agency is doing beyond your own website.

The question to ask: “Which questions are we tracking, how often are we showing up compared with three months ago, and what are you doing to improve that?”

When anyone can write with AI, how do you tell if your content marketing is any good?

So you’re publishing aggressively. Blogs go up every week, social posts are scheduled, and the agency sends you a content calendar that looks impressively full.

However, you can’t quite see what all these leads up to. Where are the leads from your content marketing efforts exactly? 

Content is the discipline most often measured by output instead of outcome. Agencies count posts published. There’s more to this than just the volume. Why? Because we live in the age of AI and it brings its own challenges. 

The AI problem and the abundance of volume 

Up until a few years ago, writing a good blog post took real time and skill and original research. Today, anyone can produce a thousand words in thirty seconds. That’s changed the landscape entirely. Volume is now cheap, so volume alone proves nothing.

I am not saying you should look for an agency that doesn’t use AI. We use it, and so does nearly everyone else. But if you’ve heard of ‘AI slop’ you’ll know that’s what you should be wary of. The problem with AI slop is that it’s generic, unedited and interchangeable. 

Here’s a simple test: if a competitor could publish the same piece by swapping in their logo, it isn’t doing much for you.

Content is still the king, but a very specific kind. The authentic kind of content. Stuff that only you can offer: your data, your client stories, your founder’s opinions, even the unpopular ones. 

Watch where your agency gets it’s material from. A good answer involves interviewing you or your team regularly. A worrying answer is “we research the topic.”

What to measure, by format

  • Ghostwriting and founder LinkedIn. Ignore likes from strangers. Watch profile visits, connection requests from your target buyers, meaningful comments and DMs, and prospects who say “I’ve been following your posts.” Expect this to build over months, not weeks.
  • Company LinkedIn and social. Follower count is a vanity metric. Look at engagement from the right people, clicks to your website, and whether social shows up as a source in your leads.
  • Email newsletters. Open rates have become unreliable because some email apps register opens automatically. Focus on click rate, replies, unsubscribes and how many subscribers go on to book a call or buy.
  • Blogs and long-form articles. Measure organic traffic by search intent (we already covered this in the SEO section), time on page, and whether readers go on to enquire. Some posts convert directly; most build trust over several visits.
  • White papers and reports. For many early-stage businesses, these aren’t worth the effort yet. Where they do make sense, measure how many qualified leads download them and whether your sales team actually uses them.

The question to ask: “Which pieces of content brought us the most enquiries or qualified traffic in the last three months, and what made them different from the rest?”

How do you know if your website is doing its job?

A new website is expensive and takes so much time to get right. And once it’s live, most founders breathe a sigh of relief and move on.

Don’t.

A website isn’t a finished product; it’s a business tool, and the only question is whether it’s doing its job.

Start with a baseline

The biggest red flag in any website project is the absence of a ‘before’. If nobody recorded your old site’s enquiry numbers, conversion rate and load speed before launch, there’s no way to prove the new one is better. If you’re about to start a redesign, insist on this. If you’ve already launched without one, start measuring today so you at least have a baseline for the next six months.

What a good website should do

  • Turn visitors into enquiries. Every form, call button and WhatsApp click should be tracked as a conversion in Google Analytics. If it isn’t tracked, it can’t be measured.
  • Load fast on a phone. Much of your traffic, especially in India and the Middle East, comes from mobile phones on patchy networks. Run your site through Google’s free PageSpeed Insights tool. It shows how your site performs for real visitors on Google’s Core Web Vitals, and whether it passes.
  • Be readable by search engines and AI tools. A beautiful site that Google can’t understand is an expensive brochure. Ask whether it was built with proper heading structure, clean code, schema markup and descriptive page titles.
  • Keep people engaged. Google Analytics now reports “engagement rate” rather than leaning on bounce rate. Look at which pages hold attention and which pages people leave immediately.

What to ask for 30 and 90 days after launch

  1. How many enquiries has the site generated, and how does that compare with before?
  2. What’s our conversion rate, overall and on key pages?
  3. Does the site pass Core Web Vitals on mobile?
  4. Which pages do people spend the most time on, and where do they drop off?
  5. What is our page load speed on mobile? (Should be under 3 seconds)
  6. What is our bounce rate, and how does it compare to before?
  7. How many enquiries or leads has the website generated this month?
  8. Which pages are people spending the most time on?

How do you measure the success of product marketing?

Product marketing is the least understood service on this list, so let me explain what it covers. It’s the work that decides what you say about your product, to whom, and how it reaches the market. Everything else, from ads to content to your website, depends on it.

It usually includes:

  • Positioning and messaging. Who your product is for, what problem it solves, and why you rather than the alternatives.
  • Go-to-market (GTM) strategy. The plan for launching a product, feature or new market: audience, channels, pricing, timing and what success looks like.
  • Launches. Coordinating everything that goes out when something new ships.
  • Sales enablement. Pitch decks, one-pagers, competitor comparisons and objection handling that help your sales team close.
  • Customer and competitor research. Understanding why people buy, why they don’t, and how competitors position themselves.

How to measure it, realistically

Product marketing is hard to measure because it works through other people. Good positioning shows up in better ads, better sales calls and a better website, rarely as a number with its own name. That’s not an excuse for having no numbers, though. Agree on them before the work starts.

How to measure it, realistically

Product marketing is hard to measure because it works through other people. Good positioning shows up in better ads, better sales calls and a better website, rarely as a number with its own name. That’s not an excuse for having no numbers, though. Agree on them before the work starts.

  • For positioning and messaging: win rate against key competitors, sales cycle length, and conversion rates on pages using the new messaging. Message testing through ads or landing pages can give you early signals within weeks.
  • For a launch: set targets before launch day. Sign-ups, demos booked, pipeline created, or feature adoption among existing customers within 30, 60 and 90 days.
  • For sales enablement: does your sales team actually use the materials? Ask them. Unused decks are a cost, not an asset.
  • For research: did it change a decision? Research that sits in a folder hasn’t done its job.

Be patient with the lagging numbers. Revenue and win rates take a quarter or two to reflect a positioning change. But you should see leading signals much sooner: sharper sales conversations, better-qualified leads and clearer feedback from prospects.

The question to ask: “What does success look like for this launch or positioning work in 90 days, and how will we measure it?”

What should you track in Google Ads, and what should you ignore?

Google Ad are the easiest to track, and still easy to dress up. I’ll explain. 

aid search is the most measurable channel on this list. You spend money, people click, and you can see what happens next. Results should show within weeks, not months. Campaigns do need a short learning period (usually two to four weeks), but after that you have clear visibility on what’s happening. 

Unfortunately though, it stop ads reports from being padded with numbers that look good and mean little.

The vanity metrics in ads reports

  • Impressions and clicks. More of both just means you spent more.
  • Click-through rate on its own. A high CTR from the wrong audience is an expensive way to get irrelevant visitors.
  • Low cost per click. Cheap clicks are often cheap because nobody valuable is competing for them.
  • “Conversions” without a definition. Check what’s actually being counted. I’ve seen accounts where a page view or a button click was counted as a conversion, which makes the results look several times better than they are.
  • Brand campaign results. Ads on your own brand name almost always perform brilliantly, because those people were already looking for you. They’re worth running, but they shouldn’t carry the report.

The numbers that matter

  • Cost per qualified lead or cost per acquisition, based on leads your team has actually checked.
  • Return on ad spend from your own sales data, not just the figure the ads platform reports about itself.
  • Branded versus non-branded performance, reported separately so you can see whether ads are bringing in new demand.
  • The search terms report. This shows what people actually typed before clicking your ad. Read it yourself once a month as it’s often the fastest way to spot wasted spend.

Red flags and good practice

  • You should own your Google Ads account. The agency should have access to it, not the other way round. If you leave, your data and history should stay with you.
  • Ask about negative keywords. These stop your ads from showing on irrelevant searches. A well-run account adds them regularly.
  • Check that ads lead to relevant pages. Sending every click to your homepage wastes money.
  • Connect ads to real outcomes. The best setups feed back which leads became customers, so campaigns are optimised for sales, not form fills.

The question to ask: “How much did we spend on non-branded campaigns last month, how many qualified leads did that produce, and what did each one cost?”

The simplest test of all

If you takeaway only one thing from this blog post, make it this question:

“What did marketing do for our business this month?”

A good agency can answer it in plain language, in a couple of minutes, without a deck. Everything else in their report is supporting detail.

Here’s a cheat sheet to save or share with your team:

ServiceWhen to expect real signalsWhat to measureWatch out for
Google Ads2–4 weeksCost per qualified lead, non-branded results, ROAS from your sales dataVague “conversions,” brand campaigns carrying the report
Website30–90 days after launchEnquiries, conversion rate, mobile speedNo “before” numbers
Product marketingWeeks for leading signals; 1–2 quarters for revenueWin rate, launch targets, sales team usageNo success criteria agreed upfront
Content3–6 monthsQualified traffic, enquiries per piece, engagement from buyersOutput counts, generic AI copy
SEO3–6 months, longer in competitive categoriesRankings and traffic by funnel stage, organic enquiriesTraffic growth from irrelevant keywords
GEO and AEOTrends over several monthsMention rate across repeated prompts, AI referral trafficGuarantees, judging from a single check

One more thing: hold up your end

As we wrap this piece up, I just wanted to share a final reminder: measurement works both ways. An agency can only show results against goals you’ve actually agreed on. 

So share your sales data, give access to your analytics, agree on the numbers that matter before the work starts, and resist changing the brief every month. 

The best client relationships I’ve seen weren’t the ones with the biggest budgets. The best, always, were the ones where both sides were looking at the same numbers.

And if your agency is doing good work, tell them. Give them a shoutout or make a referral before they ask. It matters more than you’d think.

Want to see what an honest report looks like?

At leatherjacket studio, every report we send starts with that one simple question: what did marketing do for your business this month? If you’re unsure whether your current marketing is working, we’re happy to take an honest look, even if the answer is that your current agency is doing a great job.

Write to us at hello@leatherjacket.studio.

Back to Blog