Part of our guide hub: Digital marketing
Measuring marketing is straightforward when someone clicks an ad and buys within the hour. It becomes genuinely hard when the purchase takes four months, involves three people, and the first touch was a blog post nobody remembers reading.
Most B2B and considered purchase businesses are in the second situation, and most of them measure as if they were in the first.
Why last click attribution misleads
Default analytics credits the last thing someone clicked before converting. In a long cycle that is almost always a branded search: they typed your company name because they already knew about you.
So the report says branded search produces everything and the blog produces nothing, when the blog is what made them type your name. Businesses act on this and cut the channel that was working.
The honest position: with a long sales cycle and a small budget, perfect attribution is not achievable. Accepting that is more useful than buying a tool that produces confident numbers built on assumptions.
The two questions that beat any dashboard
1. Ask every new customer how they found you
One question, asked by the salesperson, recorded in the CRM. Not a dropdown on a form, which people select at random, but a human question in the first conversation.
The answers are messier than analytics and closer to the truth. “I read your article on ERP costs six months ago, then saw you on LinkedIn, then asked a friend who used you.” No attribution model produces that, and it tells you exactly what to keep doing.
2. Ask what they read before deciding
Asked after signing. It identifies the specific pages that carry weight in the decision, which are frequently not the ones with the most traffic.
Twenty of these conversations tell you more about your marketing than a year of dashboards.
What to actually track
Four numbers, maintained honestly, beat an elaborate system nobody updates.
- Qualified enquiries per month, by source. Qualified meaning they could plausibly buy, not every form submission.
- Cost per qualified enquiry, by channel. Total spend on that channel divided by qualified enquiries from it.
- Win rate by source. The number that reveals which cheap channel produces expensive time wasting. A source with a low cost per enquiry and a five percent win rate is worse than an expensive one closing at forty.
- Average deal value by source. Referrals and content commonly produce larger deals than paid search, which changes the ranking entirely once you look at revenue rather than lead count.
Multiply the last three together and you get the number that matters: revenue per rupee spent, by channel. That is marketing ROI, and it takes a spreadsheet rather than a platform.
The measurement setup a small team can maintain
- UTM tags on every campaign link, consistently. Agree a naming convention and write it down, because inconsistent tagging is what makes reports unusable a year later.
- A source field in the CRM filled by the salesperson from the human question, not auto populated.
- Analytics with conversion tracking for the digital half, understanding that it undercounts.
- Call tracking if phone is a real channel, which it is for most Pakistani businesses.
- One monthly review where the four numbers above are updated and looked at together.
That last point is the whole thing. Businesses with mediocre tracking and a disciplined monthly review make better decisions than businesses with sophisticated tooling and no review.
Leading indicators worth watching
With a four month cycle, waiting for revenue to judge a change means acting a quarter late. Useful earlier signals:
- Branded search volume. Rising means awareness work is landing, and it moves before revenue does.
- Direct traffic. People typing your name.
- Returning visitors before enquiry. Most considered purchases involve several visits.
- Enquiry quality, judged by sales rather than by count.
The trap of measuring only what is measurable
Channels that report cleanly, such as paid search, get credited and funded. Channels that report badly, such as content, referrals, events and word of mouth, get cut because they cannot prove themselves.
This is a systematic bias and it damages businesses. For most service companies in Pakistan, referrals and reputation produce the largest and best deals, and neither appears in an attribution report.
The defence is the human question. Ask everyone, record it, and let the answers correct what the dashboard says.
What good looks like
You should be able to say, in a sentence: we spent X last quarter, produced Y qualified enquiries at Z each, closed W of them worth V, and the channels that produced the best revenue per rupee were these two.
If you cannot say that, the problem is usually not the tooling. It is that nobody is recording the source at the point of sale.
For the channel comparison itself, our breakdown by cost per qualified lead ranks the realistic options for a services business.
Frequently asked questions
What is a good marketing ROI?
It depends on margin and customer lifetime value rather than on a universal number. The useful comparison is between your own channels, and against the cost of the sales time each one consumes.
How do I attribute a sale that took six months?
Imperfectly. Use the human question at the point of sale as your primary source, and treat analytics as a supporting view rather than the truth.
Is content marketing measurable?
Partially. Direct attribution understates it because it usually appears early in the cycle. Ask new clients what they read; that is the most reliable available measurement.
Do I need an attribution platform?
Not below substantial spend. UTM discipline, a source field in the CRM and a monthly review cover most of what a small business needs.
Ezitech runs marketing for service businesses and builds the analytics behind it. See our marketing services or talk to us about measurement.