How to use marketing attribution to grow your sales pipeline

Written by Keith Holloway | Oct 27, 2023 12:13:00 PM

Do you know marketing activities drove all your new leads and revenue?

Are you confident the data is accurate? Are you sure?

If yes, then congratulations, you are a rare breed.

But guess what?

It might not be as accurate as you think. I’m willing to bet there are plenty of reasons it just isn’t so.

Why?

Tracking is hard. There are so many potential points of failure. The truth is it’s almost impossible to get the 100% true answer to attribution.

But you can get very close.

And certainly close enough to get solid direction on how to allocate your marketing budgets for more success.

In this article I’ll cover the most common types of tracking used by marketers today, and the associated problems, what it takes to get to full-funnel tracking, and the benefits of going down this path:

Why is great marketing attribution so difficult to solve?

Before we dive into specific problems and how to address them, let’s take a look at some of the most common setups for marketing tracking or attribution and their shortcomings.

Online-only tracking

This category of tracking is most popular because it’s easiest. Solutions include Google Analytics, Google Tag Manager, all the tracking provided by media platforms such as Google Ads, Facebook and LinkedIn.

Even though it’s the easiest, there are still many potential points of failure.

Some of which you can control:

Incorrectly configured UTMs, pixels, events and goals
Missing tracking codes on pages
UTMs on internal website links, or HTTP links on a secure site
GDPR configuration

And many you can can’t

Offline touches
Do not track settings
Multiple devices
Cleared cookies
Technology limitations
Word of mouth

When configured properly and well understood, it’s possible to get decent attribution to conversions but it doesn’t help you “close the loop” to opportunities and revenue.

Online-only tracking is appropriate for e-commerce companies where the entire buying journey takes place online. It’s often not even helpful for SaaS businesses, because the sale often takes place in a back-end system and is not connected to the original visits of your prospects.

For companies that have a sales department to close deals, it usually leads to poor marketing decisions since there is no connection between anonymous website conversions and your sales pipeline.

A connection between your website traffic and your CRM is required for complete attribution.

Using landing pages for attribution

We’ve seen many companies attempt to use landing pages to identify the source of leads. Many marketers believe this is an easy and failsafe way to identify lead sources.

This couldn’t be farther from the truth.

The idea is you create a landing page for Google, another for LinkedIn, another for Facebook, etc., and then you send all traffic from those platforms to those pages.

Sounds like a simple and obvious solution to know where your leads came from, right?

There are a few major problems with this approach:

It's “last touch” tracking
As previously mentioned, with a “considered” sale there are usually many touches that precede a lead. With this method, you are only measuring the leads that came directly from that campaign, on the same visit. But how many additional leads were touched by that campaign and later become leads through another source?
It presumes people will not break out of the landing page
Even if your landing page doesn’t have a single link to your website, people will leave the page and go to your site directly. People are smart. They can just type in the domain directly even if you avoid linking it because they want to check out the full site.
We’ve done the testing. We’ve seen as many as 40% of leads sent to a landing page found their way onto the main website and submitted a form there, not on the landing page.
There is no granularity in campaigns or keywords
Unless you are prepared to build out a landing page for every single campaign, you are only measuring channels and not campaigns. Individually campaigns within a channel will perform very differently. And that’s not to mention individual keywords.

All things considered, using landing pages to measure marketing attribution is a non-starter.

We’re doing attribution to understand the true impact of marketing, right?

When using landing pages for marketing tracking, you are selling your marketing team short because you only capture a small percentage of the leads generated or touched by your campaigns.

Online-only tracking connected to your CRM

The obvious next step for your sales-driven company is to connect website sources to your CRM.

Many marketing automation solutions, like HubSpot and Salesforce Pardot, include some type of marketing attribution that will identify the source of leads and allow you to connect your sales pipeline to original website sources.

Of course, your results will depend entirely on how well the tool is configured, although with a tool like HubSpot, this is relatively straightforward.

Challenges appear when connecting your marketing automation platform to a CRM like Salesforce.

Most Salesforce-centric organizations measure revenue on the CRM side, and original sources are often lost or not carried through the integration.

Furthermore, and most difficult to solve is taking into account all the marketing or sales touches that do not happen on your website:

Leads generated in third-party media platforms such as LinkedIn or Facebook Lead generation campaigns
Webinars hosted by third parties where the leads are imported to your CRM and have no website touch (or at least no connection to their website touches)
Webinars hosted by yourself, but where a third-party form, such as Zoom, was used to capture registrants, that circumvent the website tracking you have in place
Any imported list, such as leads generated at a tradeshow or contacts that were added to the CRM directly, such as through an inbound phone call
Outbound sales activities that touch contacts in your database and are not connected with website sources.

When there are many offline and online touchpoints in play, marketing attribution can easily be at odds with the sales team.

Sales takes credit for leads that were originally generated by marketing.
Marketing takes credit for leads that were originally generated by sales.

To solve these problems, we need to be tracking the entire ecosystem.

Full-path tracking

The ultimate form of marketing attribution considers all online and offline touchpoints from prospects and even related company contacts in generating pipeline and revenue.

This is much more difficult to solve because you need a connection to every website session and every touch point measured in the CRM. There are many points of failure and virtually every problem identified in this article needs to be addressed to solve for accuracy.

While challenging, the results are more than worth it, because it's the only way to get to the truth of what’s working.

Generally, if your marketing budget is greater than $5K/mo the results will pay for itself. If your marketing budget is $100K/mo or more, it’s a no brainer.

What would happen if you had accurate marketing attribution?

Getting attribution right can save you thousands in wasted marketing and make a material impact to your pipeline and revenue.

And even knowing it’s just directionally correct, the knowledge of what’s working and what isn’t will make all the difference in allocating your marketing budgets.

How would it improve your marketing decisions, your business outcomes, or even your career?

1. You’d stop wasting money on conversions that don’t lead to business

I recently read a story about a marketer that spent $1 million on LinkedIn advertising and generated thousands of leads, only to learn much later that only a handful of the leads turned into qualified opportunities and absolutely none went to closed/won. A million dollars in leads didn’t sell one $2,000 product!

For a year, that marketer was celebrated for bringing in so many leads at such a low cost.

Until the truth was discovered through attribution. Ouch.

2. You’ll be investing in campaigns that generate pipeline and revenue

Oftentimes, the activities that bring in the lowest cost per lead have the highest cost per opportunity and lowest revenue generated compared to the investment.

Without attribution, you could be ramping up the worst campaigns and throttling the best.

With full-funnel attribution you will avoid measuring the effectiveness of marketing campaigns by marketing conversions and cost per conversion.

And instead measure effectiveness by sales pipeline generated, the cost to generate that pipeline, and the associated sales conversions or revenue.

Full Funnel tracking
Measuring the pipeline and revenue generated against the investment into
the channels and campaigns to drive that pipeline.

3. You will have more pipeline and revenue for same dollars invested

Once you identify the waste and fully-fund the channels and campaigns that work, there will be a material impact on your sales pipeline and revenue.

4. You’ll be taken more seriously from your CFO, CEO and Board

Only with full-funnel attribution can you understand the impact of marketing investments, make a sound business case for investment and be taken seriously by your CFO and board.

What do you need to do to solve for full-path attribution?

Sold? OK, so let’s talk about what to do.

Get your online tracking in order

Getting your online tracking configured correctly is the first step to achieving some semblance of data accuracy.

In addition to incorrectly tracking online touch points, many online tracking problems will affect measurement further down the funnel.

Plan your UTMs well, avoid using non-standard values, and execute on your UTM plan consistently
Since most paid advertising measurement is dependent on these values, having well planned and consistent values is key.

Thousands of hours have been wasted by marketing analysts trying to figure out the mess made here by trying to match up the related campaigns.

Hopefully they got it right in their analysis, but it’s obviously better to just do it right in the first place.

What’s a UTM?

UTM parameters, those ubiquitous ?utm_medium/source/campaign values you see at the end of advertising URLs. They’ve become a defacto standard of web measurement since Google bought Urchin Software in 2005 and relaunched it as the free product known as Google Analytics today. Fun fact: UTM stands for Urchin Tracking Mechanism and is an artifact of that purchase.

Ensure your pixels are firing on the correct events
I can’t count how many times we’ve seen campaigns generating incredible results, only to find out that the pixels or goals are tracking how many people saw the page, instead of submitting the form, which the goal was ostensibly measuring. The reverse can also be true, a campaign could be delivering results, but not tracking to it.
Make sure your tracking codes on all pages
Missing GTM, Google Analytics, HubSpot, insert your tool here, code on some pages will play havoc on your tracking. For best results, use a website crawler to check every page.
Ensure you don’t have any UTMs coded on internal website links
Another common mistake is to add UTMs to internal pages. We’ve seen marketers or developers add utms on an internal link to track internal website actions, such as “CTA-from-Blog” or “pricing-page-demo”.

Not only are there better, built-in ways to measure this, what you don’t realize is every time someone clicks a link with these parameters a new website session is recorded and the original session information is lost. There aren’t many better ways to ensure you are undercounting your paid advertising results!
Make sure you don’t have any http links on your https site
Similar to adding UTMs to internal links, when a visitor clicks an internal HTTP link and then gets redirected to the correct HTTPS version of the page, the original visit source is lost and you now have an internal referral. Once again, reducing the measured impact of your efforts.
Setup cross-domain tracking and use referral exclusions where appropriate
If you have two domains that cross link to each other, you’ll probably want to know how people originally found your sites, not that one of your sites lead to a visit to the other.

A related problem is using an OATH or payment gateway on your trial or purchase flows. We frequently see the most important conversions on sites being attributed to Google, Visa, Paypal or other systems when the visitor didn’t actually come from those sites.

Instead they were simply redirected in the background to that site and then returned to your site during their conversion process.

This is addressed by adding referral exclusions, to look up the previous session start and ignoring these temporary referrals that are starting a new session.

Take steps to close the loop and implement full-path attribution

Since these next points are critical to getting to full-funnel attribution, I’ll go into these in greater detail:

Connect your leads with original sources
Connect your opportunities with the original leads
Branch out beyond a single type of attribution
GOTCHA: Your entire Total Addressable Market (TAM) is already in your database

Connect your leads with original sources

This is the most obvious first step towards marketing attribution for lead generation.

Online marketing platforms revolve around tracking conversions.

That’s why most agencies and marketing departments put all their efforts on tracking conversions.

But you know what? A conversion does not equal a qualified lead!

But why?

It’s an invalid/fake submission
It’s a competitor
It’s a contact already in your database, submitting a form again
It was a test
It was a real person, but not a lead in your target market

The only way to address these problems is to connect your digital marketing with your CRM with marketing attribution software so that you know which leads were generated from your campaigns.

Connect your opportunities with the original leads

Once you are tracking the source of your leads, the biggest win is to connect opportunities (deals) in with those leads.

This is straightforward in theory, but often breaks down in practice.

For example, in Salesforce, you can “convert” a lead to a contact, company and opportunity.

As long as this is done correctly, the lead will be connected with the opportunity and you will be able to track the original source of the opportunity.

What can go wrong? Here’s a few, not uncommon, scenarios:

A prospect submits a form and then calls. Sales takes the call and enters a new contact, unconnected with the lead form.
The person that submitted a form on the website is not the person sales engages with the initial sales call. A new contact is created and connected to a new opportunity that is disconnected from the original lead.

The solutions here are varied as there are multiple things that can go wrong when human processes are involved.

Practical solutions include:

an SLA between marketing and sales
regularly auditing of opportunities to ensure they have been properly created and connected

Some companies, with very high-volumes of leads, have gone down the path of implementing “fuzzy matching” technologies to match leads and opportunities (although we’re not sure how much we’d trust the accuracy of that).

Branch out beyond a single type of attribution

Prospects visit your website and have many touches with your business before they engage with you – especially in high-value B2B markets.

There are many ways to measure this and no one method provides all the marketing guidance you need. For example:

Attribution Model	What it means	What to watch out for
Last touch	The most recent way the prospect came to your website. It is usually referred to as the method your prospect found your website immediately before submitting a form on your website. This is the most basic form of attribution and is the easiest to measure, since you only need to identify the more recent website session.	It tends to overcount branded and direct touches and undercounts the channels that introduce potential new customers to your business. Different platforms measure this in different ways so it may not even mean what you think. (I’m looking at you Hubspot, with your “latest touch” reporting, which doesn’t mean what most people think it does)
Converting touch	This is a more nuanced and more accurate name for last touch. It literally means the source of the session that resulted in the conversion.	The same issues apply for last touch, but at least it means what you think it means.
First touch	The very first measured touch of your website, which could be days, weeks or months before they submitted a form Generally the best method for measuring lead generation campaigns because it measures how visitors first found out about your company.	This is more difficult to measure since you need to connect all previous website visits to each form submission. All of the tracking gotchas listed in a previous section can break this chain – multiple browsers, deleting cookies, Do Not Track, etc. However, despite tracking difficulties, this method gives very strong signals as to what campaigns are most effective. An important caveat: Remarketing campaigns, by definition, will not show many results on a first touch model since all contacts touched by remarketing should already have been touched by marketing in some other campaign. This is one reason why other methods are required:
Evenly weighted	All previous website touches are attributed equally to the conversion.	Since it’s a middle-ground between first and last/converting touch, this will provide some credit to all touches so it does provide some credit to campaigns in the middle of the buyer journey, such as email, remarketing etc. However, typically we want to bring in as many new prospects as possible, so you run the risk of underweighting the budget on the campaigns that introduce new visitors.
U-Shaped	All website touches are taken into account, but more weight is applied to the first and converting touches	This might be a bit better than evenly-weighted since it applies more weight to the first touch, but it also waters down the results by heavily weighting the last.
Time Decay	Applies the most weight to the most recent touch, and trails off on weighting to older touches.	This is often used for very long sales cycles, and makes sense on the surface since you don’t want a campaign from 4 years ago to take credit for a sale today. However, it will overly emphasize the most recent touches, so once again, risks over budgeting on marketing to people who already know you.
Custom	Custom models applying weights specific to your situation	While this may seem to be the ideal solution for those who have the tools, skills and time to tweak models, it avoids the biggest issue. No one attribution model can give you the best answer to every question.
First paid touch	Provides credit to the first paid campaign a contact interacted with	This is a model we came up with at PureSEM to identify the total impact of paid marketing campaigns. This model is especially helpful when you have a relatively small TAM (total addressable market) and most of your potential customers are already in your database. However, it will overemphasize the importance of branded search campaigns, so like all models, it should be used in conjunction with others to get the answers you need to make decisions.
First web touch	Provides credit to the first website touch of leads, assuming you are tracking both online and offline touches.	This is a handy model to identify the true impact of digital marketing campaigns when you have a large database. The challenge is it requires you to be measuring both online and offline touches with complete date sequencing built into a consolidated marketing data warehouse containing “all touches”.

As you can see, different models are better at answering different questions.

For example, if you want to know how your remarketing campaigns are impacting results, you’ll want to look at models like Last/converting touch, equal weighted, or first paid touch.

If you want to know what’s introducing new customers into your marketing funnel, the best model is first touch.

Gotcha: Most of your TAM is already in your database

Marketing Attribution Models like those in HubSpot break down when most of your target market is already in your database.

Real example:

More than one of our clients have a large database of contacts that contains most of the potential buyers in their market.

In the CRM, all of these contacts will have been imported so they have an original source of “offline sources”.

The reality is, most of these contacts have never interacted with the business.

So what?

When any of these contacts engage with a marketing campaign the first touch (original source) of each of these contacts will remain “offline sources” on a first touch basis, no matter what the engagement.

So even with first touch, you will completely miss the impact of the marketing activities on lead generation!

In this case, first paid touch, and first web touch are great choices, followed by equal weighted or even time decay.

A solution to increase pipeline with marketing attribution

At PureSEM, we’re obsessively focused on investing in marketing that drives revenue.

So much so, we’ve put our money where our mouth is and built a full-path marketing attribution software system (currently in beta, invite only) specifically designed to increase the pipeline generated from digital marketing campaigns.

We are taking into account all touches in the sales and marketing funnel, not just web touches, and not just the original contact, but all contacts related to your deals.

Furthermore, since we’re primarily focused on search engine marketing, we think we have a pretty unique and valuable set of tools to measure the effectiveness of both organic and paid search built right in.

Have questions or comments about attribution? Have a problem you’d like to solve? Drop us a line to chat, or request a demo to learn more about our software and services.

View full post