Customer Analytics

Fix Your Funnel – The Very, Very Top of It!

Share this post:

In my last two posts, I focused on analyzing and optimizing your funnel once visitors had gotten to your website. But there is an even bigger opportunity to ‘widen the top of the funnel:’ the ability to optimize your marketing programs so you get even more visitors to visit your website in the first place!

The Last Click Problem

A well known limitation of some analytics programs is the ‘last click’ problem. Historically web analytics credited all purchases and conversions to one click, the last click.

Years ago, Coremetrics (now IBM Digital Analytics), realized that the last click problem was a big problem. They realized that visitors don’t just come to your site one time and convert or purchase. No, they need to do research, to compare prices or compare you against your competition. And even when they’ve found exactly what they wanted, they may need to leave your site one more time and do a search to see if they can find a coupon. Because of this behavior, Digital Analytics was built with Attribution baked right into most reports.


What is Attribution?

Attribution is understanding conversion credit beyond the last click. Marketing performance metrics can be calculated based on looking back for a period of time, and using attribution settings in the calculation.  Clients can customize the amount of time that used to lookback based on their actual latency. 

Some common attribution settings are:

    • First Click – Giving only credit to the marketing program that brought the visitor to your website initially
    • Last Click – Giving only credit to the marketing program that brought the visitor to your website when the conversion or purchase took place
    • Even or Average Distribution – Giving credit to all of the clicks from all of the marketing programs in the Customer’s Journey
    • Custom Algorithm – Some clients decide they want to give a higher percentage to the first click (maybe 60%), less credit to the last click (maybe 10%) and divide the remaining credit (30%) to any other click that occurred in the Customer’s Journey

And because IBM Digital Analytics allows you to configure multiple settings, you can compare the performance of your marketing programs with these different views.


Marketing programs that have a strong ROI when measured by First Click are the programs that do a good job of introducing new visitors (and buyers) to your digital property. Marketing programs that have a strong ROI when measured by Last Click do a good job a ‘closing’ the sales. Marketing programs that have a strong ROI measured by Average or Even Distribution, help move a visitor along the sales process.

Attribution Matters

Making decisions with bad data – meaning incomplete or incorrect – always has consequences. Making marketing decisions, what to run, what to buy, what to stop doing, based on last click data isn’t just silly – it leads you into making poor decisions.

When you only credit the last click, you place greater, value on clicks from marketing programs that are at the bottom of the funnel, the programs focused on the visitors who already know you. Programs like Branded Keyword PPC Campaigns, Affiliate or Deal-Site clicks perform extremely well when measured by Last Click. Programs like non-branded keywords or display often don’t perform as well when measured by Last Click. Because of this, decisions can be made to stop spending money on general, non-brand terms, not invest in display, and focus only on branded keywords and email.

The problem these type of decisions create, is they kill the top of the funnel. Sure, it may seem great that you can be highly focused on branded keywords and email, and this may even free up budget to move to other initiatives. In the short term, conversions probably look good and ROI probably goes through the roof. For a little while, you are the hero.

But then it happens. Sales drop off sharply. Conversions based on this focused group of visitors who are already aware of your brand eventually dry up. And the damage has been done. Stopping the types of marketing that introduces new visitors to you and your website has killed the ‘top of the funnel.’


You can see why it is very important to be able to give credit to all of the clicks that result in a purchase or conversion, to accurately understand the performance of all your marketing channels. When you start understanding this full customer journey, you will likely make different decisions.

Over the years, when a company goes from measuring and making decisions on last click data, to moving towards crediting the full customer journey, some of the consistent outcomes I’ve seen are:

    • They become able to justify a higher cost-per-click for non-brand terms
    • They gain an understanding of the branding value of display (especially when giving credit to both impressions and clicks with Impression Attribution)
    • The find that Social does play a strong role in the Customer Journey, sometimes at the very beginning, but more often in the middle of the journey
    • They realize they may have been over-crediting (and therefore over paying) their affiliate partners
    • They see the average order value (AOV) and sales per customer increase dramatically when a visitor interacts with multiple marketing channels

And the lessons go on and on. So when you think about fixing your funnel, don’t just start when visitors get to your website. Make sure to start at the very top of the funnel, with what is driving visitors to your site.

For a better understanding of marketing attribution capabilities with IBM Digital Marketing Optimization, feel free to check out this white paper. Happy attributing!

More stories

Innovating with AI and Blockchain for Professionals

We recently announced a definitive agreement under which HCL Technologies intends to acquire a select number of IBM marketing and commerce software products. HCL is a valued IBM partner with whom we already have a development partnership with, in support of the products in the announcement. HCL and IBM expect to close the transaction in […]

Continue reading

A New Era for Marketing and Commerce

Last week, we announced a sharpened focus on AI-powered and cloud-based applications, and that HCL plans to acquire certain marketing and commerce software. You can read more about the announcement here. While the transition is not planned to take place until mid-2019, I thought it was important to share what this means for our portfolio […]

Continue reading

How to Turn Customer Loyalty Programs into Love

“Loyalty programs increasingly need to have flexibility. In other words, it can’t just be ‘Buy ten get one free,’ at a coffee shop anymore. Stores themselves are starting to have to get creative with what loyalty means to them and giving the customers choice in what those rewards are.” — Nicole Leinbach Reyhle, Founder, @RetailMinded […]

Continue reading