Digital Marketing Examples: 8 Channels With the Numbers

Most “digital marketing examples” articles show you a famous brand’s campaign, which is entertaining and useless — you do not have their budget, their audience or their agency. These are eight examples of the kind of work an actual small business or client account involves, each with the arithmetic written out, so you can substitute your own numbers and see whether the channel works for you.

Every figure below is an assumption, labelled as one. The point is the shape of the calculation, not the numbers.

1. Organic search — a local service business

The example. A plumbing firm ranking for emergency searches in its town.

Assume Search Console reports 900 impressions a month for its priority queries at an average position of 6, with a 5% click-through rate — so 45 clicks. Assume 12% of those clicks call.

900 × 0.05 = 45 clicks, and 45 × 0.12 = 5.4 calls a month.

Now assume the work lifts click-through to 10% at the same impression count: 900 × 0.10 = 90 clicks, and 10.8 calls. The number of impressions never changed. Doubling the outcome came from the title, the description and the position — not from publishing more pages, which is what most SEO plans start with.

When it is the wrong channel. When there is no existing search demand for what you sell. Search captures demand; it does not create it. Check the impression count before committing a year to it, and if the demand is there, decide which half of search to fund first — margin and sales cycle answer that faster than a preference for organic does.

The file: SEO tutorial for beginners walks the sequence through the site’s working files rather than explaining SEO in the abstract.

2. Paid search — a B2B services firm

The example. A $3,000 monthly budget on high-intent search terms.

Assume a $4 cost per click, a 4% landing page conversion rate, a 25% close rate and an average deal of $2,000.

• Clicks: $3,000 ÷ $4 = 750
• Leads: 750 × 0.04 = 30
• Cost per lead: $3,000 ÷ 30 = $100
• Sales: 30 × 0.25 = 7.5
• Revenue: 7.5 × $2,000 = $15,000, a ROAS of 5.0×

Where it breaks. Halve the conversion rate to 2% and the cost per lead doubles to $200 with nothing changing in the ad account. Paid search is the channel where the landing page, not the bidding, usually decides the outcome.

The file: the PPC audit checklist has 28 checks and a workbook that gives a negative/watch/keep verdict per search term, plus the recoverable total.

3. Paid social — an ecommerce brand

The example. $2,000 a month on prospecting for a $120 average order value.

Assume a $12 CPM, a 1% click-through rate and a 1.5% purchase rate.

• Impressions: $2,000 ÷ $12 × 1,000 = 166,667
• Clicks: 166,667 × 0.01 = 1,667, an effective CPC of $1.20
• Orders: 1,667 × 0.015 = 25
• CPA: $2,000 ÷ 25 = $80
• Revenue: 25 × $120 = $3,000, a ROAS of 1.5×

When it is right. When you are creating demand rather than capturing it, and the margin can carry the acquisition cost. Work out your break-even ROAS before the first campaign, not after the first invoice — the same campaign run through ROAS, ROI, CPA and CAC shows how far apart those four answers land when the margin is included.

4. Email — the same ecommerce brand

The example. An 8,000-person list, one send a week.

Assume a 2.5% click rate on the list, a 3% conversion rate on those clicks, and the same $120 average order value.

• Clicks per send: 8,000 × 0.025 = 200
• Orders per send: 200 × 0.03 = 6
• Revenue per send: 6 × $120 = $720
• Monthly: $720 × 4 = $2,880, against a platform cost in the tens of dollars

This is the cheapest revenue in the list, and it only exists because a previous channel paid to build the list. Email is not an acquisition channel — it is what makes acquisition worth doing twice.

Where it breaks. Deliverability. A list that does not reach the inbox has all the same numbers and none of the revenue, and open rate will not tell you. Under Apple’s Mail Privacy Protection, “remote content is privately downloaded in the background when you receive a message (instead of when you view it)” — Apple’s own documentation — so an open can be a machine fetching a tracking pixel for someone who never looked. The email marketing checklist uses bounce and complaint rate for exactly this reason, and has no open-rate column at all.

5. Content marketing — a professional services firm

The example. Twelve articles over a year, six hours each including research and editing, at an internal cost of $60 an hour.

12 × 6 × $60 = $4,320 of effort in year one.

Assume three of the twelve reach 200 sessions a month and the other nine reach almost nobody. That is 600 sessions a month at a 2% enquiry rate — 12 enquiries a month, arriving from work that was already paid for.

The two things that matter in that calculation are both uncomfortable. The payback is deferred by months, so content cannot fix a quarter that is already going badly. And nine of twelve did nothing — which is normal, and the reason to plan a year of content rather than one article.

The file: the content plan template is one row per piece, with a refresh date and a counter for how many are planned, published and due a rewrite. Working out which three earned the sessions is a different job — the content audit template returns keep, update, consolidate, redirect or delete per page from your own clicks, impressions and conversions, which is the calculation that stops the other nine being repeated next year.

6. Organic social — the same firm

The example. 22 posts a month at 25 minutes each, including the thinking.

22 × 25 = 550 minutes, about 9 hours, or 9 × $60 = $540 a month of unpriced time.

Assume that produces three enquiries. Cost per enquiry: $540 ÷ 3 = $180 — against $100 in the paid search example above, where the cost was at least visible.

Organic social is rarely free. It is a media spend paid in hours, and it should be compared with the alternatives on those terms. It is worth doing when the audience genuinely gathers there, or when the proof of expertise is the product.

The file: the social media audit template computes cost per enquiry per channel and gives a keep-or-close verdict, which is the calculation almost nobody runs.

7. Conversion rate optimisation — the B2B site from example 2

The example. 4,000 sessions a month converting at 2.2%.

• Now: 4,000 × 0.022 = 88 conversions
• After a change taking it to 2.7%: 4,000 × 0.027 = 108 conversions

Twenty extra conversions a month with no additional media spend. At the $100 cost per lead from example 2, that is the equivalent of $2,000 a month of paid search, bought once.

8. AI search — measuring rather than guessing

The example. Forty prompts a customer might actually ask an assistant, checked on a schedule, with whether your site was cited recorded each time.

Assume you are cited in 6 of the 40 — a 15% citation rate. If a competitor is cited in 22 of the same 40, that gap is a measurement you can act on, and it is worth more than any ranking-factor theory.

The mechanism is partly documented, and it is worth being precise about which part. Google’s guide to generative AI features on Search says they are “rooted in our core Search ranking and quality systems”, describes grounding as retrieval through those same systems, and defines query fan-out as “a set of concurrent, related queries generated by the model”. So the plumbing is public. What is not published is the weighting — no provider gives an ordered list of factors or a formula you could reproduce, which is why a confident ranking-factor claim is still an inference from a small sample.

That same guide is unusually direct about what you do not need: “You don’t need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search”, structured data “isn’t required for generative AI search”, and there is “no requirement to break your content into tiny pieces for AI to better understand it”.

Which leaves measurement as the actual work. Measure citation rate, note which competitor was cited instead, and fix the pages that should have been quoted. The AI search page includes a tracker that does this across assistants.

What to do with these

Two things.

Compare cost per outcome, not cost. Example 2 costs $3,000 a month and example 6 costs “nothing”; one produces leads at $100 and the other at $180. Until the hours are priced, the comparison is not being made. The same blind spot applies to the tools these channels run on: setup and maintenance are a third of what a stack costs in its first year and almost never appear next to the subscription.

Work backwards, not forwards. Every example above starts from a spend and calculates an outcome, which is the wrong direction for planning. Start from the outcome you need and derive the traffic — the funnel calculator does this across all six stages and makes the binding constraint obvious, which is usually reach rather than the landing page everyone wants to optimise.

Then choose which channels to fund and which to explicitly not fund. The strategy examples show four businesses doing exactly that, including the channels each one ruled out and why. That choice is the strategy; the dated version of it with owners and costs against each line is a different document with a different reader, and producing one when the other was asked for is a week gone. If it is one campaign rather than a whole strategy, the campaign template has the brief.