7 Steps to Build Marketing Attribution

Don’t Build an Attribution Monster …

This is where B2B tech startups need to be particularly careful. It is possible to build an extraordinary attribution system. You can connect the CRM to marketing automation, website analytics, advertising platforms, social platforms, content systems, event platforms, intent data and sales engagement tools. You can introduce first-touch, last-touch, linear, U-shaped, W-shaped, time-decay and predictive models. Then, you can build dashboards, calculate attributed revenue, construct customer journey maps, create scoring systems and spend six months maintaining the whole thing. Meanwhile, nobody is creating enough content, engaging prospects, running enough campaigns or talking to customers. The attribution system becomes the work rather than a tool for improving the work, which is completely backwards.

Marketing attribution must serve marketing. Marketing must not serve attribution.

Focus on What You Need to Understand

Attribution is about helping you better understand what you need to know about your go-to-market activities.

You want to know things like:

• Where are our leads coming from?
• Which channels are producing opportunities?
• Which campaigns are creating engagement?
• Which activities appear repeatedly in successful customer journeys?
• Which activities produce pipeline?
• Which activities produce customers?
• How long does it take to move from first identifiable interaction to opportunity and then to revenue?

Those questions alone can provide enormous value, and you don’t need a PhD in statistics to begin answering them.

7 Steps to Starting Your Marketing Attribution

Step 1: Define What You Want to Know

Before starting and certainly before buying attribution technology, decide what business decisions the data needs to support. You may want to know whether to spend more on events or whether content is producing pipeline. You may want to compare paid advertising with organic demand generation, understand whether your sales development activity is generating incremental opportunities or establish whether marketing is contributing enough pipeline to justify its budget. Write those questions down because they become the foundation of your measurement strategy.

Step 2: Establish a Small Number of Meaningful Funnel Stages

Don’t create twenty-seven lifecycle stages because your CRM allows you to. Start with something everyone understands.

For example:

• Target account → Engaged account → Lead → Qualified opportunity → Closed won

The precise terminology doesn’t matter nearly as much as everyone agreeing what each stage means.

Marketing and sales must use the same definitions, because otherwise the attribution model is simply measuring disagreement.

Step 3: Capture the Source and Campaign Properly

Make sure every campaign has consistent tracking and use the following where appropriate:

• UTM parameters.
• Campaign naming conventions.
• Forms that capture original source information.
• Sales activity is logged.
• Opportunities are associated with the right contacts and accounts.

Don’t underestimate data hygiene, because an attribution system built on poor data will produce beautifully presented nonsense.

Step 4: Start With Simple Attribution

Use first-touch and last-touch reporting initially, not because either is perfect, but because they are easy to understand. First-touch can help answer: “What introduced us to this prospect?” Last-touch can help answer: “What immediately preceded the conversion?” Neither tells the whole story, but together they provide useful bookends while your team develops a more sophisticated understanding of its customer journeys.

Step 5: Add Influenced Attribution

The next step is to recognise that marketing can influence opportunities that it didn’t originally source. This is particularly important in B2B technology. For example, sales may have sourced the opportunity, but the prospect may have subsequently consumed five pieces of content, attended an event and watched a product demonstration. That activity matters because it influences the lead along each stage of the sales pipeline. It shouldn’t necessarily be credited with causing the deal, but it should be visible as part of the journey. This is where the distinction between sourced and influenced revenue becomes useful.

Step 6: Look for Patterns Rather Than Pretending to Know Causality

This may be the most important principle of all. Don’t ask: “Did this blog article cause the sale?” Instead, ask: “How frequently does this type of content appear in successful customer journeys?” That is a much more useful question. If 100 opportunities are analysed and 70 of the eventual customers consumed technical content before engaging with sales, that is interesting. If customers who attended events have a higher win rate, that is significant. If opportunities exposed to case studies progress faster, that is vital information. These are patterns, and patterns become evidence. Evidence becomes intelligence, and this intelligence can eventually improve investment decisions.

Step 7: Don’t Confuse Correlation with Causation

This is where attribution needs humility. Suppose customers who read your blog have a higher average deal value; does that mean your blog created the additional revenue? Maybe, or those customers could be more engaged buyers. Perhaps larger companies naturally consume more content, or your sales team directs better prospects towards your best content. There are multiple possible explanations, but the job of marketing analytics isn’t to manufacture certainty, as that is where attribution runs into trouble. What you are really doing is trying to reduce uncertainty, which is a much more realistic objective.

Technology Can Make Attribution Easier

Fortunately, marketers no longer have to build everything manually, as CRM and marketing platforms such as HubSpot now provide attribution functionality across contacts, deals and revenue, depending on the product level. HubSpot’s current attribution tools allow marketers to analyse assets, interactions, UTM data, campaigns and revenue across different attribution models.

Dedicated B2B attribution platforms have also emerged, such as:

• Dreamdata: Built specifically around B2B customer journeys, connecting marketing and revenue data and providing multi-touch attribution and journey analysis.

• HockeyStack: Provides multi-touch attribution, journey analytics and data integration across marketing and sales systems, including the ability to compare different attribution models.

• Factors.ai: Focuses heavily on B2B attribution, account-level engagement, campaign performance and connecting marketing activity with pipeline.

• Ruler Analytics: Takes another approach, particularly around visitor-level tracking, forms, calls, live chat and connecting conversions back to CRM and marketing data.

These tools are not magic; they still depend on data quality, identity resolution, sensible definitions and appropriate configuration. However, they can dramatically reduce the manual work involved in creating a useful measurement framework, making the difference between a stillborn effort and a valuable investment.

Don’t Build What You Can’t Maintain

For a small startup, this is critical advice. A £50,000 attribution project that requires a full-time RevOps specialist may be completely inappropriate for a business with £2 million in ARR. The technology and effort must match the maturity of your business. If the team consists of a founder, one salesperson and one marketer, a sophisticated enterprise attribution platform may be overkill. A properly configured CRM, clean campaign tracking and a simple monthly dashboard might be more valuable. As the business grows, the sophistication can grow with it.

What Startups Must Do Before Starting Attribution

There are several simple things you can do today that will make future attribution much easier.

• Clean the CRM: Make sure contacts, companies, opportunities and lifecycle stages are consistently recorded. Poor CRM hygiene becomes a major barrier when attribution eventually arrives.

• Standardise Campaign Naming: Don’t allow five different names for the same campaign. Establish a simple naming convention that everyone follows.

• Use Consistent Tracking: Use UTM parameters and campaign identifiers consistently across digital campaigns. The objective is to make future analysis possible without reconstructing history.

• Record Sales Activity: Marketing shouldn’t attempt to understand the customer journey without sales data. Calls, meetings, opportunities, objections and outcomes all add context.

• Connect Contacts to Companies: B2B marketing increasingly needs account-level visibility. Knowing that three people from the same target account interacted with your company can be much more useful than treating them as three unrelated leads.

• Record Revenue: Attribution ultimately becomes much more valuable when marketing can connect activity to opportunities, closed-won deals and revenue rather than stopping at lead generation.

• Don’t Delete History: Keep campaign and source information wherever possible. Historical data becomes increasingly valuable as the company builds its evidence base.

Learn and Adjust as You Go

The biggest mistake a startup can make is waiting until its attribution model is perfect before starting, because it never will be.

You cannot predict the future, but you can almost guarantee the following:

• The business will change.
• The product will change.
• The market will change.
• Buyer behaviour will change.
• The technology stack will change.
• The channels will change.
• Your customers will change.

Your attribution model will therefore need to adapt and change as well.

Think of attribution as a journey through increasing levels of confidence. At the beginning, you might have only basic source data, but then you add campaign tracking, opportunity data, influenced activity, account-level engagement and eventually you may have enough historical data to start testing more sophisticated models. That is how you measure progress.

The Real Objective is Better Decisions

This is the point that can easily become lost. The purpose of attribution isn’t to create an impressive dashboard, assign every marketing activity a percentage, prove that marketing deserves credit for every deal or calculate an absolutely accurate ROI for every campaign.

The purpose is to help marketing make better decisions, so you can answer questions like: do we:

• Invest more in events?
• Produce more technical content?
• Reduce paid advertising?
• Invest more in customer advocacy?
• Build more analyst relationships?
• Target a different audience?
• Change our messaging?
• Spend more money on brand?

These are the questions attribution should eventually help answer, and sometimes the answer will be: “We don’t know yet.” That is perfectly acceptable. A useful measurement system should be able to say what it knows, what it thinks it knows and what it cannot yet determine.

Start the Journey

Attribution is about continually pushing forward while simultaneously developing a degree of clarity to the purpose and outcomes of your actions. It’s not about spending all your time looking in the rear-view mirror. For B2B tech startups, there is a natural tension between measurement and momentum. You need enough intelligence to avoid driving blind, but you cannot spend so much time measuring the road behind that you stop moving forward.

Attribution is valuable, but it is not a substitute for marketing and without taking action you are starving the attribution of the data it needs.

Remember, attribution doesn’t:

• Create demand.
• Write the content.
• Build relationships.
• Speak to customers.
• Create a brand.
• Generate sales conversations.

It’s marketing that does all those things. Attribution is designed to make marketing progressively smarter.

The process is:

• Start with the basics.
• Capture the data.
• Keep it clean.
• Agree the definitions.
• Measure the obvious.
• Look for patterns.
• Refine and optimise.

Resist the temptation to claim certainty where none exists. Then gradually make the model more sophisticated as the organisation accumulates more data and experience. The goal is to become progressively less wrong over time, not to achieve perfection. For a B2B tech startup, that is a much more realistic and valuable definition of success.

You may want to read: “Why Tech Startups Struggle with Marketing Attribution.”