Measuring Marketing Success: From First Click to Closed Deal

Why do your marketing numbers say nothing about orders?

The agency’s monthly report shows 4,000 visitors, a click-through rate and the cost per click. Your accounting shows orders and revenue. There is no connection between the two reports. You know what you spent, and you know what you earned. What the one has to do with the other, you do not know.

This is not a special case. Shashi and colleagues (2025) reviewed 96 academic articles on the digitalization of business-to-business marketing. They find that many companies in this field do not yet base their decisions on data, and they highlight the role of marketing analytics and decision tools.

The reason is rarely missing software. The numbers arise in two separate worlds. Web analytics ends at the submitted form. The sales list begins with an email in the inbox that no longer shows where it came from. In between, the most important piece of information is lost: the source. This piece asks: how does a small business measure which marketing leads to orders? To answer it, we first introduce the measurement chain with its four points, then choose the necessary metrics and tools, assess the question of attribution, and work through an example of which channel pays off. At the end comes a 30-minute monthly review.

A marketing number that ends at the form measures effort. Only the number that ends at the order measures success.

The measurement chain: at which four points do you need to measure?

I call the path from the first click to the order the measurement chain. It has four links. At each link there is exactly one number, one place where it arises and one person responsible. The chain only holds if all four links are connected.

A worked example for one quarter shows what the chain looks like in numbers.

What connects the links is the source. It is set at the first contact and never overwritten afterwards. In practice the chain almost always breaks between the second and the third link, which is where marketing hands over to sales. How to regulate that point is described in the article on the lead handoff from marketing to sales.

Which metrics are enough for a small business?

The choice of metrics in digital marketing is vast. Saura, Palos-Sánchez and Cerdá Suárez (2017) compiled and organized the most important ones in a systematic literature review. Their aim was a shared understanding of which metrics companies need to know and manage. Above all, the work shows one thing: without a selection you drown in numbers.

Five metrics are enough to steer by, per source and per quarter. They follow the measurement chain and end at the money.

The contribution margin is what remains of the order value after deducting the directly attributable costs. If you do not know it, start with the order value and a cautious discount. The metrics for sales itself, such as proposal turnaround and order size, are covered in the article on sales metrics for small and midsize businesses.

Click-through rate, time on page and reach remain useful for troubleshooting. If the inquiry rate drops, they help you understand why. They do not belong in the monthly meeting with management.

Which tools are enough?

The measurement chain needs four building blocks. None of them is expensive, and most businesses already have three in house.

Tagged links

A tagged link carries small additions at the end that tell web analytics where a visitor comes from. The usual ones are the so-called UTM parameters: utm_source for the source, utm_medium for the type of channel and utm_campaign for the campaign. Consistent spelling is important. If you write “Newsletter” once, “newsletter” once and “NL” once, you end up with three sources.

The source field in the form

The most important step is storing the source together with the inquiry. Technically, the form reads the tag from the link and writes it into an invisible field. That way the source arrives in the email or the CRM that sales works in. Always add the open question of how the prospect found you. It captures referrals, trade shows and conversations that no web analytics can see.

1. Define sources

   Write a list of no more than ten sources, for example search engine, search ads, LinkedIn, newsletter, referral, trade show.

2. Tag links

   Every link in ads, newsletters and profiles gets the matching tag according to fixed naming rules.

3. Store the source

   The form passes the source into a hidden field. Inquiries by phone get their source by hand.

4. Add the outcome

   Sales enters conversation, proposal, order or rejection in the same row.

The setup is craft work and done in a few days. If you do not want to do it yourself, you can outsource it, for instance to quik-services, a provider of websites, SEO and lead generation and a business partner of Owners Council. The same test question applies to every agency: does it report up to the inquiry or up to the order? Further questions are listed in the article on choosing an agency.

Attribution: which channel does the order belong to?

Attribution is the assignment of an order to the channels that contributed to it. That sounds simple, and it is not. A customer reads an expert article, sees you weeks later at a trade show and finally inquires through a search ad. Which channel brought in the order?

It is documented that the choice of rule changes the result. Li, Kannan, Viswanathan and Pani (2016) analyzed six months of data on several hundred search keywords at an online jewelry retailer. In the middle of that period the firm switched from last-click to first-click attribution. The calculated returns of the search keywords differed significantly depending on the rule. For this firm, first-click attribution led to lower returns, most markedly for very specific keywords. The authors also explain why the rule matters so much: it defines which keywords count as effective and therefore which ones the firm bids higher on in future. If you only count the last click, you easily cut back on the general keywords through which customers find the provider in the first place. If you only count the first, you can overrate them. This is a single retailer in the consumer business, but the point is general: the rule determines where the budget goes.

A second paper shows the limits of the common method. Gordon, Moakler and Zettelmeyer (2026) analyzed 2,226 advertising experiments at Meta, where one of the authors works. In such experiments a randomly chosen control group does not see the advertising, so the actual incremental effect can be measured. The industry-standard last-click attribution within seven days explained only about one fifth of the differences in incremental effect per dollar spent. A prediction method that learns from the experiments reached just under nine tenths. That does not make the last click worthless: in a simple decision model by the authors it led to a different decision than the experiment in 12 to 20 percent of campaigns, the prediction method in 8 to 12 percent. For a small business such experiments are hardly feasible. The lesson is useful all the same: the number in the ad account is an approximation and not proof.

My suggestion for a start: record the first touch as the source and ask in the first conversation how the customer found you. Where the two differ, note both. That is rough but honest, and it is enough for budget decisions in a midsize business.

Worked example: which channel pays off?

With the inquiry list you can calculate after one quarter, for each source, what one euro spent brought back. In the example, every order yields a contribution margin of 5,000 euros. The costs include your own working time, valued at a fixed hourly rate.

The example makes no statement about which channel is better. It shows the form of the analysis. With numbers this small, chance plays a part: one order more or less shifts the ranking. So read the table across several quarters and do not cut a channel because of one weak quarter. Channels such as expert articles also work with a delay. What is realistic there is covered in the article on SEO for midsize businesses.

The monthly review in 30 minutes

Numbers only change something when someone looks at them regularly and decides. One fixed appointment a month with marketing and sales is enough. The agenda always stays the same.

1. Look at the measurement chain per source: visitors, inquiries, conversations, orders.
2. Name the point with the biggest loss.
3. Decide on one single change for the next month.
4. Check whether last month’s change worked.

More than one change per month makes the effect unreadable. If you shorten the form, change the ad and cut the response time all at once, you will not know afterwards what helped. How to reach sound decisions with a few numbers is explored further in the essay on data-based decisions.

How to get started

Start with what you already have. The measurement chain is built fastest backwards: from the order to the source.

• List the orders of the last six months and note for each where the customer came from. Ask if in doubt.
• Create a list of no more than ten sources with fixed spellings.
• Add the question “How did you hear about us?” to the inquiry form.
• Set up the hidden source field and the tagged links with your agency or web provider.
• Create the columns source, outcome and order value in the inquiry list.
• Schedule the monthly 30-minute appointment.

Even the first exercise, sorting the orders of the last six months by origin, is revealing. If you have to ask about several orders, you have found the first gap in your measurement chain. How marketing and sales work together with these numbers is described in the article on aligning marketing and sales.

Back to the opening question of which marketing leads to orders: it can be measured if every inquiry keeps its source from the first click to the close. A web analytics tool, tagged links, a source field in the form, and a shared list are enough for that. Every attribution of orders to channels remains an approximation. For the decision, however, rough but honest numbers are enough, and they are preferable to precise but misleading ones.