Visualizing the Marketing Funnel

Howdy, SEOmoz fans. Welcome to another edition of Whiteboard Friday. This week we’re talking about the inbound marketing funnel. Specifically, we’re talking about why basic web analytics can often fail and mislead marketers in to investing in a lot of the wrong channels. What I have done today is try to illustrate that visually with this beautiful funnel. I know, I know, it looks a little complex. You’re going to go, “Boy, that is really colorful and you must have taken a long time to draw those lines.” I did.

But think about things this way. Let’s imagine the flow of visitors to your website. What happens is there are some visitors who are coming to you for the first time. Never seen your site before, you’re sort of capturing them early in the stage. Maybe they know very little about you. Maybe they’ve heard something about you through word of mouth. Maybe they saw someone tweet a link. Maybe they found you through a search. Maybe they found you through just typing in your web address directly or through an email someone sent them. Whatever the case might be, you can track all of those. You know where they come from. I’ve simplified this. Obviously, there are more channels and you could break these up into direct, search, social, RSS, your blog, and referring links. Referring links from sort of the blogosphere, the news world, forums, links on other corporate sites, whatever it is. There could be advertising things in here too, but I am going to ignore paid for the moment.

What’s interesting is when you think about this, think about the people who come to you for the first time and how you capture them. Those channels might be entirely different from the people who way down here at the bottom of this funnel completed a transaction that was worth money. So these are people who made me directly dollars or Euros or whatever . . . Euros. Is that the Euros symbol? I don’t know. Let’s do the yen symbol. I know that one better. Made you cold hard cash. Excellent. Great.

But what happens? Marketers almost always, and particularly the higher up you go in a marketing organization, they look at what sent traffic here and that’s where they assign the budget. Right? Because they’re tracking actions down here and they go, “Oh, well, fantastic. You know, these people down here, look at this. Social in green sent no completed transactions, screw social. Stop wasting your time on Twitter and Facebook and LinkedIn and Google+. Those are a waste of time, because look, they didn’t send us any transactions.” Oh yeah? How did all these people know to subscribe to your RSS feed? How did they know to come from referring links? How did they know to type in your web address directly? How did they know to search for you? Particularly if this is branded and unbranded search, if I break that our right here, I bet this is going to be 90%+ branded search, meaning people who searched for your brand name or items related to your brand name, not unbranded terms that they’ve never heard of before and aren’t familiar with. Because of this, you attribute revenue and people and time to places that they shouldn’t go, and this is a very, very dangerous thing.

So let’s walk through the funnel. I think that everyone who has Google Analytics or something more advanced installed can build this out. In fact, I built this for SEOmoz’s own website recently. I’ll give Kenny a screenshot to put in the Whiteboard Friday of that specific image that I built. It’s not very advanced, but it can give you a sense of what ours sort of looks like. So, you can take a count of the first-time visitors. So, it’s like, oh, I’ve got a million first time visitors in the last 60 days, and maybe I have 750,000 returning visitors in the last 60 days. How many people made 4+ visits in the last 60 days? Well, that’s interesting. You can segment. You can create a segment that’s says, “Hey, only show me people who have visited 4 or more times in the last 60 days, and now show me the channels that they came through. Show me the referring sources of those individuals.” Wow. Now I can start to see a pattern emerging. I can see, oh look at this. This was actually true for SEOmoz. RSS very little at first time, bigger at referring, and then a nice segment, a really nice segment, right down to completed transaction where people don’t really come through RSS very much when they complete a transaction.

You can do this. Let’s say there’s 300,000 who came in the last 60 days 4 or more times, and maybe there’s 100,000 who came 10+ times. Wow. These people are really interesting. Particularly these people are really interesting because they must really like what you have. They must like your content or your blog or something that you do on your site, some sort of tool, some sort of function that you provide to them. Yet, there’s going to be a very different look at who sent these visitors, right, which of these sources sent these visitors versus which sent the people who made a conversion action, something like signed up for an email account, or signed up with their email address, or subscribed to your newsletter, or filled out a survey, or filled out a request for a white paper that gave you a lead versus actually completed a transaction, like went back to the site and signed up for the paid service or bought the product, or did whatever it is that makes you money. Knowing these source differences and building out this funnel, visualizing this funnel and being able to see these means that you will be able to do a bunch of things right.

So, I have some action items that I want you to take away from this inbound marketing funnel visualization.

Number one, please, if you can do nothing else, at least set up first touch attribution. So you may not be able to get the concrete segment that has made at least one visit from this source before converting somewhere in their funnel, because a lot of those people who convert down here are going to be in the 10+ or 4+ visit segment, and they’re going to have come to you multiple times, and usually you’re only doing last touch attribution, meaning you only see the source that sent them the last time. There are some more advanced analytics like Mixpanel or KISSmetrics that can help you see deeper into that multitouch funnel, but at least you can set up first touch attribution tracking in Google Analytics. There is a good blog post on how to do that Will Critchlow from Distilled wrote. We’ll link you over to that.

Second step, segment the content that is driving people, so not just the visit sources, but the content that’s driving people early on in your funnel and later in your funnel. Understanding that dichotomy will give you a sense of, hey, we can’t just give up on the content that’s bringing people in here or the content that’s bringing people in the 4+, 10+, made a converting action. We need to focus on both of those, and here’s how we should be distributing our time. Let’s not spend, you know, be careful not to over assign value with resources of any kind, with dollars, with people, with time, to channels that are just getting you into the bottom of the funnel. It’s really, really dangerous. It can mean that what happens over time is you increase conversion rate here, yeah, and things are going well, but as this dries up, your competitors are taking this traffic. They are getting it one way or another. Or you’re not executing on it, which means no one is happy and finding what they need on the Web from you or whatever it is that you provide.

Finally, the other thing is track those actions that are likely to lead to transactions. So, if you know that a high percentage of people in a certain bucket, in a number of visits bucket, in a number of visits to specific content bucket, in a conversion-like action, say they did this thing like sign up for an email newsletter, track the percent and the number that are flowing down to actual dollar value transactions. If you do that, you’ll have a great sense of where you can invest sort of in the middle of the funnel that will help to drive that action further down.

This is a complex fascinating process, and I know it’s not easy to implement. But for those of you who do, the returns can be phenomenal.

Thanks very much, and I hope I’ll see you again next week for another edition of Whiteboard Friday. Take care.