I remember the frustration gathering funnel numbers at the digital agency I worked at in 2012. I just couldn’t tell the whole story. Pieces were missing and data was incomplete. My second attempt at building a marketing funnel went much more smoothly at my second startup, thanks to good tool choices and starting small.
I’m going to describe in a few hundred words what probably took 2–3 months of trial-and-error and an all-but-busted Delete key on my Mac. I hope this blog helps a few other marketers shave off weeks’ worth of busywork — but this will be a complex undertaking for anyone. Set your expectations.
Choosing the right tools (and friends)
I’d always start with tools. I think I spent the bulk of my time setting up my funnel by tweaking data sources in tools like Google Analytics (“GA”) and the tag manager. Verify, verify, verify every data point.
Don’t try to do it alone, either; I have an amazing support network of people and I hope everyone could be as lucky. I leaned heavily on the expertise from friends at other startups, infinitely smarter in data wizardry than I am, to help me verify data and tweak settings. If you’re stuck, all you have to do is ask.
Here is the list of tools that currently makes up my marketing data stack:
Google Analytics
This is where top-of-funnel data on sessions and marketing site behaviour lives. It’s also something of a hub for other tools to merge data into.
Google AdWords
We spend a significant budget on our paid keywords trying to attract customers at the buying stage. Integrating AdWords into GA is easy to do — and you can track user conversions all the way through. It’s often solid, insightful data on customers.
Intercom
Love these guys! Intercom lets us stay close to the customer — a very big part of startup culture. Our support team will remedy tickets using IC and our sales & marketing teams share data on users.
ChartMogul
I usually check our MRR growth as soon as I’ve got my morning coffee. Connect Stripe to track customer payments.
Microsoft Excel
To be honest I’ve migrated a lot of my spreadsheet work to Google Sheets for the sharing features. Either way, you’ll be flexing your data muscles here.
Identify the data that you need
Start bottom-up and always with objectives.
Bottom-up metrics
If you’re visualizing a funnel, you’ll usually picture the revenue stage at the bottom and the wider user acquisition stages at the top like an inverted pyramid:
That makes sense from a buyer’s journey point of view. However, when planning marketing activities and understanding how to influence behaviour it’s important to start at revenue and work up.
I’ll say this right up front to squash doubt: it’s totally okay to be wrong now. This exercise is necessary to create the structure of your funnel. You can triage data later. For now, complete this with what you’ve got.
Take this dummy data marketing funnel as an example:
- The revenue (MRR) goal for this company is $1,000.
- At the company’s seat price of $55 per month, they’ll need 19 new customers to meet that goal.
- How do they attract 19 new customers? Look through historical data for a rough conversion rate. This company estimates they convert one out of every four free trial users into a paying customer. That means they need 76 free trial users to meet their goals.
- Same concept for trial users. Looking at historical website data tells them that every month they convert ~4% of traffic into trial users. That math leads them to 1,900 unique site visitors per month to generate the right number of leads.
- Finally this company learns that in order to earn that volume of traffic to their site, they need to be loud in the market to the tune of 126k impressions. That’s the total audience that is exposed to their brand — and believe me it’s a very rough estimate.
The very definition of an impression varies depending on where you look. Just know that it’s the sum of all ads, blog reads, social messages and outbound content from a brand. Understand that each of these channels will convert at a different rate, and it’s up to marketers to decide where to focus their efforts.
Create objectives then focus activities
Great, you’ve got a skeleton funnel based on your best-guess conversion data. It’s a starting point — now paint your future. Decide which areas need improvement and make them the focus of your activities this period.
Every single number in that funnel can be influenced by certain marketing activities.
- Generally, rate of conversion (%) numbers are influenced by optimizing activities: subject line testing, website on-page optimization, onboarding improvements, reducing friction.
- Volume numbers (#) are most easily influenced by frequency. The number of emails sent, the content release schedule, the ad budget, social updates.
- Finding the right balance between optimizing and messaging is something only your brand can answer.
Plant some goals next to every funnel metric you want to influence. Can you easily increase your impressions by updating more frequently? Set that goal. Call it 250k impressions next month. Can you convert a higher percentage of trial users into paid with a discount program? Awesome — plant a flag at 35% paid conversion and focus your activities to meet it.
Benchmarks
This data means very little until you compare it to a proper benchmark. The best benchmark is your own month-over-month comparison, which we’ll get to shortly, but for now I’ll suggest some rule-of-thumb benchmarks you might want to keep in mind as you set goals or explain your funnel to others. (Please spend some time looking outside for more relevant benchmarks before relying on these.)
- Converting impressions into site visits is sifting for gold. If you’re able to catch the attention of 2% of your audience you’ve made an impact.
- Visitors to your site are at the very least curious about your offering. Tempting anywhere from 2–10% of those into taking an action is a good use of time. Consider form completions, adding items to the shopping cart, requests for quotes and other non-paid actions as conversions.
- Finally, the number of users who actually get out their wallet and put their money where their mouse is (I’m hilarious I know) can be a little higher because they’ve already committed some sort of currency of time/info already. Aim to see 10–35% of converted users follow up on their first conversion and pay for your offering.
Organizing your funnel spreadsheet
The metrics I use to evaluate marketing performance is pretty evident from the spreadsheet I use, pictured below. Hard to miss those big green fields on the right; acquisition rate, conversion rate, paid rate. That’s just me. I like being efficient with my efforts and I’ll try to squeeze out as many users at the next stage as possible. I dislike funnel drop-off and this spreadsheet is a great way to identify where the highest fall-off is.
You don’t have to follow this model. Whatever story you want to tell your stakeholders is up to you — this particular design is great for a general view of all channels and their change month-over-month.
- If you’re a content marketer, for example, you’d probably want to focus on pageviews and engagement, maybe even time on page.
- If you manage social communities then you’d want deeper insight into engagement, time-of-day stats and sentiment.
- PPC or SEO folks will want numbers around keyword net change, spend, CPC, CTR and more in the top of the funnel.
How to calculate month-over-month change (MoM%)
The most important thing to demonstrate is how these numbers change over time thanks to the efforts of your marketing team.
- Once you’re satisfied with your formatting, duplicate the sheet and rename it for the following month. In this example I’ll use Dec 2015 and Jan 2016.
- Add in all your historical data for both months.
- In January’s MoM% column you can use this formula to calculate the difference from December (where [this is a cell] );
For net change: =([Jan 2016 value]-[Dec 2015 value] / [Dec 2015 value])
Returning a value of a few percentage points net change. Eg. 4.5% or -4.5%.
For overall change =([Jan 2016 value] / [Dec 2015 value])
Returning the value of the quotient. Eg. 104.5% or 95.5%.
How do you get December 2015’s value? When you’re typing a formula in January’s cell using Excel or Sheets, you can switch sheets mid-entry and click on the December cell with the value you want. The program will reference that cell on a different sheet for the calculation.
A little further reading
There are smart people who do this for a living. The coolest of them would be, in my opinion, Dave McClure who managed to incorporate pirates into the business of marketing; and Eric Ries who will tell you the only metrics that matter are those you can learn from.
Leave me questions and suggestions in the comments. If I can clarify or improve this post, your input would be very helpful. Thanks.
