Would you rather optimize the path your visitor will actually take or optimize the path you think he should take?
If you’d rather the former, then there’s something you need to know about linear funnels… they’re not a completely accurate representation of reality. The question is, what should you do about it and what is an accurate representation of reality?
You have to learn to optimize for the modern funnel, which is similar to a tornado. [Tweet It!]
Table of contents
- What are linear funnels?
- Linear Funnels vs. Tornadoes: What’s the Difference?
- Are linear funnels a thing of the Past?
- What Makes Tornadoes More Difficult?
- How do you manage tornadoes?
What are linear funnels?
Linear funnels are traditional processes that define each step in an ideal customer lifecycle, where the user moves from the top of the funnel to the bottom progressively and predictably.
Linear Funnels vs. Tornadoes: What’s the Difference?
Let’s look at the difference between these types of funnels. For example, someone searches for “car insurance” and finds a PPC ad…
After clicking it, they’re taken to the landing page where they click the most dominant call to action…
They’re then taken to a lead capture form, where they fill out all of the information…
Later, they’re contacted by a sales representative and sold on the idea of choosing that specific car insurance company.
The problem is that reality is much more complex and, well, messy. At eMetrics Summit in San Francisco, Josh Aberant, VP of Growth at SparkPost, went so far as to say that linear funnels are dead…
Josh Aberant, SparkPost:
- “Funnels have been blown to bits
- Users don’t take linear paths to and through apps or products
- They get to your app or product how they want to, when they want to and do whatever they want to do
- Users own the relationship”
Instead, he introduced the concept of tornadoes, which are much less organized, more complex and less predictable.
For example, someone searches for local car insurance and finds thousands of options, but focuses on the top three results first…
All three pages are open in the browser for comparison…
The person compares the landing pages, trying to determine which is the right option for them…
What about motorcycle insurance? Can I just get insurance directly from the insurance companies? What are other people saying about this company? All of these questions lead to pages like this…
Josh suggests that actual customer journeys are much more complex than a linear funnel would indicate.
After the pages above, the visitor might talk to colleagues at work to get their advice and end up purchasing 3 days later. She might turn to social media, look at more car insurance providers, go into the office, read online reviews, sign up for a newsletter first, etc.
Note that tornadoes can extend beyond the purchase. It’s not exclusively about what happens at the top of the funnel. That same visitor might have a bad customer service experience and tell her friends. She might start looking into other brokers, forget to make a payment, etc.
Here’s a good way to visualize the difference between a linear funnel and a tornado…
Awareness -> Evaluation -> Purchase -> Usage -> Repurchase -> Advocacy is an example of a linear funnel. All of the dots associated with each step of the funnel create a tornado.
Are linear funnels a thing of the Past?
Both Harvard Business Review and Practical Ecommerce have reported the death of the linear funnel. Is it true, are linear funnels a thing of the past?
Sujan Patel of ContentMarketer.io thinks we’ve gone from 2D to 3D funnels because the decision making process has become more complex…
Sujan Patel, ContentMarketer.io:
“I still think funnels are linear, but more and more factors are part of the decision making process. Gone are the days of simple funnels and now there are many external factors that influence a purchase decision, such as content marketing or social media. The best way to describe this change is going from 2 dimensional to 3 dimensional funnels.
It’s only going to get more complicated and it’s important to have the right analytics software that can handle it. Heap Analytics and Mixpanel are a couple that can track these new 3 dimensional funnels.”
For Sujan, linear funnels haven’t been “blown to bits” as Josh claims. They’ve merely expanded and become more complex due to external factors that affect the visitor’s purchase decision. With the right analytics tool, you can simplify those factors and track your new, 3D funnel.
David Arnoux of Growth Tribe shares Sujan’s love for linear funnels, but admits they have a major shortcoming…
David Arnoux, Growth Tribe:
“The case against linear funnels has been around for quite some time. Like any model it has its pros and cons. The biggest shortcoming being that it considers the sales cycle to be a linear process where a customer passes sequentially through sales steps and where he/she is the only one involved in the decision making process.
The customer discovery journey (CDJ) popularized by McKinsey was the first counter-reaction to linear funnels. In the CDJ users can come in at any point, jump in, jump out. It looks more like a circle than a line.
Having said that I still love funnels. They have the huge advantage of being easy to understand. When something is easy to understand it’s simpler to communicate and implement across teams. It’s easier to understand and to adopt.”
So, in summary, the pro of the linear funnel is that it’s easy to understand and communicate, and the con is that it’s a bit disconnected from the reality of the modern customer journey.
Here’s a closer look at the customer discovery journey that David mentions…
While McKinsey talks of profound changes…
Marketers have long been aware of profound changes in the way consumers research and buy products. Yet a failure to change the focus of marketing to match that evolution has undermined the core goal of reaching customers at the moments that most influence their purchases. The shift in consumer decision making means that marketers need to adjust their spending and to view the change not as a loss of power over consumers but as an opportunity to be in the right place at the right time, giving them the information and support they need to make the right decisions.
…I’m not convinced their customer decision journey is that drastic of a departure from linear funnels (aside from the fact that it’s presented as a circle).
What is important, though, is that McKinsey focuses on “consumer-driven” marketing. Instead of waiting for you, the marketer, to drive the relationship, consumers are driving the relationship themselves. They don’t necessarily move through the funnel as you intended them to.
Morgan Brown of Inman News shares an opinion similar to David’s…
Morgan Brown, Inman News:
“Linear funnels shouldn’t be taken as as an exact replica of reality. Like any good system diagram they simplify a complex set of interactions into something that can be easily grokked and conceptualized. They can still be very helpful in understanding your essential customer flows and identifying opportunities. In many cases companies would take leaps forward just to have a super-solid understanding of this simplified view.
But of course funnels aren’t reality, it’s much more complicated than a one way slide down a series of sequential gates. The issue is we can forget to consider the greater complexity that isn’t captured by the funnel, limiting our perception of what’s really happening. When a funnel keeps us from considering the broader and more nuanced customer journey it can hurt our opportunities to find growth.
One of my favorite examples of this is the design of viral loops. Most funnels treat referrals as another step in a funnel, but of course we know they don’t work that way. Loops create interactions that lead to subsequent loops and so on. Andrew Chen has said that some of the best growth people don’t think in terms of funnels, they think in terms of viral loops. And I think this is really the point.
Funnels are helpful to a point but can be limiting if used in the wrong manner. So my recommendation would be to ensure you understand your funnels but also realize that there is more going on and keep that reality front and center with a customer journey map or similar so you don’t get inadvertently boxed in.”
So, are linear funnels an overly simplified version of reality? Yes. Does it make them a “thing of the past”? Not exactly. They can still be used to communicate growth and optimization to executives, or to foster interdepartmental understanding and goal setting, for example.
But they should not be mistaken for an accurate representation of reality, which is much more complex and messy. When they are, you miss out on crucial points of optimization and growth.
Tornadoes may not be replacing the linear funnel any time soon, but they are a good way to think about that complex, messy reality.
In the words of George E. P. Box, “All models are wrong, but some are useful.”
What Makes Tornadoes More Difficult?
As you can imagine, tornadoes are a bit more difficult to deal with than simple linear funnels. Why? There are two core reasons (among many others): growth attribution and points of optimization.
1. Dealing with growth attribution
When asked about whether he believed linear funnels were a thing of the past, Ed Fry of Inbound.org brought up the issue of growth attribution…
Ed Fry, Inbound.org:
“Linear funnels were probably never a thing. There are many different channels now – attribution is the real bitch. How do you really know how your last user or customer found you? Or all of their touch points.
This starts with getting your data in one place. For us at Inbound.org, it’s immensely powerful when we bring everything we have on an individual together in one tool so we can segment and adjust messaging appropriately (or try to… that’s another story).
Your marketing’s only as good as your data. Get all that data in one place. Different channels make it harder, sure, but it’s table stakes to compete at a high level these days.”
If linear funnels were an accurate representation of reality, we could simply know that if Customer X made it to Point B, they came from Point A. Unfortunately, it’s not that easy because, as Ed points out, having so many different acquisition channels, devices, purchase influences, etc. makes it difficult to attribute growth.
Chris Mercer of SeriouslySimpleMarketing.com explains how to address attribution in the wake of tornadoes…
Chris Mercer, SeriouslySimpleMarketing.com:
“Direct response marketers continue to use linear funnels, but they are starting to evolve as their business models move toward cultivating a relationship with their customers (and building a brand in the process).
We think of funnels in terms of main milestones (i.e. sales page, cart, purchase, etc.) The tricky part is there are many different paths from milestone to milestone. Rarely do most users complete a funnel in one session. Instead, different traffic sources work together to introduce, then nurture the visitor through a particular path within the mail funnel.
Attribution can be challenging, but you can do amazing things with Google Analytics and proper tagging of your traffic. For example, instead of tagging paid Facebook traffic as just “facebook”, instead try adding “retargeting” or “first touch” so you know what part of the Facebook campaign is working. Instead of tagging your email traffic as just “email” (something most people never actually do) try adding additional parameters so you know what type of email it was (content vs. sales message).
When you do that, your Multi-Channel Funnel reports in Google Analytics can go from a generic Social > Email > Conversion attribution to something a little more useful, like Facebook-First Touch > Facebook Retargeted > Email – Content > Email – Sales > Conversion.
With better UTM tagging you open up a more detailed view of the specific path visitors are taking to complete the various milestones in your funnel.”
If you’re not already, you should be familiar with UTM tagging. Buffer wrote a great beginner’s guide to UTM tagging, which you should take 5 minutes to read.
2. Dealing with additional optimization points
Another issue is that with tornadoes comes more optimization points. It’s not enough to be focused on “engagement”, for example, you need to focus on all of the elements that go into that step of the funnel.
Here’s David on what you need to keep in mind when optimizing modern funnels…
David Arnoux, Growth Tribe:
“The solution moving on is not necessarily to replace linear funnels but rather to improve them by including a deeper level of behavioral analysis and segmentation. Here are a few recommendations:
- Understand that there are multiple touch points needed for a customer to navigate through the different steps of the funnel. Most of the marketing community has accepted this and we see a huge surge of retargeting efforts.
- It’s not one funnel fits all. I would argue that each customer segment requires their own funnel. Each funnel will have different friction points and possibly require different UX, sales and marketing efforts. Additionally it’s no longer enough to think of segments as “customer types” (for example a B2B SaaS segmenting by small, medium and large companies). Consider that customer segments can also be a mix of technological and behavioral attributes such as device, purchasing history, purchase recency, pages visited, day of week, traffic source and so forth. Events tracking and big data have paved the way for great progress in the field of predictive analytics and behavioral segmentation.
- Understand that decision makers are often a group of people and not a sole person. It’s important to think about what the decision process is and how each player can be influenced into making the purchase. Decision makers can range from family, friends, prescribers to co-workers. Think for example of an online learning tool for kids where children, parents and teachers are all involved in the decision to adopt a tool.”
So, in summary…
- Visitors rarely make purchase decisions on their first visit, which is why retargeting and attribution are paramount. (Only 2% of visitors make a purchase on their first visit, according to AdRoll.)
- Segmentation is more important than ever because you need to understand and optimize a number of different customer paths. (We wrote an extensive article on data segmentation, which you should take a few minutes to read.)
- As the popularity of personalization rises, we may begin to see sites embracing tornadoes and creating unique, fully customized funnels for each visitor.
- Decisions are not always made by a single individual. Your value proposition has to be well-rounded and your funnel has to be accommodating.
At the end of the day, the process is still the same: conduct the research, fix the obvious issues, create hypotheses for the more complex issues, prioritize the hypotheses, run the tests.
What’s different is that you need to mentally move away from the linear funnel concept. Or, at least, move away from the idea that it’s an accurate representation of the behavior of your customers.
As Chris said, be sure to include your Multi-Channel Funnel reports in Google Analytics in your conversion research process. (If you’re unfamiliar with conversion research, you can learn more via the ResearchXL model.) Otherwise, you’re really limiting your ability to optimize effectively.
How do you manage tornadoes?
During his presentation, Josh mentioned something very important…
Josh Aberant, SparkPost:
“The classic funnel stages still matter… users just don’t go through them linearly.”
You’re not throwing years of marketing know-how out the window here. All you’re doing is recognizing that visitors don’t always move through funnels linearly anymore and adjusting your optimization process based on that knowledge.
So, how do you manage tornadoes? The same way you manage linear funnels… with a few tweaks.
1. Identify user behavior
While there are a number of different classic funnel stages, for this article, we’re going to use Acquisition -> Activation -> Engagement -> Retention -> Resurrection. Anyone who works for a SaaS company will recognize these funnel stage labels, but with a name change, they’re relevant to every industry.
The first step is to understand your users’ behavior with conversion research. There are two core questions you’re looking to answer early on:
- How do users activate and engage? These are high-value optimization points. Where are they coming from? What traits do they share? What paths did they take most often?
- What is healthy user behavior? You want to optimize for this behavior. What are the signs of a healthy, long-lasting user? What do they all have in common?
Once you know this, you’re in a better position to come up with hypotheses for creating healthy user behavior.
You’re likely familiar with Facebook’s 7 friends in 10 days metric, right? It’s called a correlative metric and a lot of different companies use them to optimize for healthy user behavior…
Stewart Butterfield, Slack:
“Based on experience of which companies stuck with us and which didn’t, we decided that any team that has exchanged 2,000 messages in its history has tried Slack—really tried it.
For a team around 50 people that means about 10 hours’ worth of messages. For a typical team of 10 people, that’s maybe a week’s worth of messages. But it hit us that, regardless of any other factor, after 2,000 messages, 93% of those customers are still using Slack today.” (via Fast Company)
Facebook has 7 friends in 10 days, Slack has 2,000 messages. How do you turn that into real results…? Andrew Chen of Uber explains…
Andrew Chen, Uber:
“Once you have a way to evaluate the success of a user, then you want to grab a cohort of users (let’s say everyone who’s joined in the last X days) and start creating rows of data for that user. Include the success metric, but also include a bunch of other stats you are tracking- maybe how many friends they have, how much content they’ve created, whether they’ve downloaded the mobile app, maybe how many comments they’ve given, or received, or anything else.
Eventually you get a row like:
success metric, biz metric 1, biz metric 2, biz metric 3, etc…
Once you have a bunch of rows, you can run a couple correlations and just see which things tend to correlate with the success metric. And obviously the whole point of this is to formulate a hypothesis in your head about what drives the success metric. The famous idea here is that, fire engines correlate with house fires, but that doesn’t mean that fire engines CAUSE house fires.” (via Quora)
Andrew talks a bit about cohort analysis, which you can learn more about here.
So, once you have that healthy user behavior metric, you can begin to isolate what correlates with that metric. Once you know that, your optimization decisions will be 10x smarter because, well, you know what to optimize for.
2. Analytics and establishing benchmarks
Next, you’ll want to set up your analytics configuration. If you’re just getting start with analytics and want to be sure you set up Google Analytics properly, read these…
- Google Analytics 101: How To Configure Google Analytics To Get Actionable Data
- Google Analytics 102: How To Set Up Goals, Segments & Events in Google Analytics
If you Google Analytics already set up, read this…
Obviously, you want to ensure your analytics setup is configured properly. If Peep were writing this, he’d tell you that most are not. Take the time right now to get it right before you continue.
Once you have everything configured properly, you’re looking to answer one question:
- How often are they currently activating and engaging? How often are healthy users activating and engaging? How often are at risk users activating and engaging? How often are users on their way out activating and engaging?
You want to answer this question early on so that (a) you can chart growth and (b) you can recognize poor user states before it’s too late.
Based on your research and analytics, you should create hypotheses to answer these three questions:
- How can you encourage healthy user behavior? Look at your healthy users. How can you create more users like that using your healthy user behavior metric? How can you keep those users healthy?
- How can you get them to activate and engage more? Look at your at risk users. How can you encourage engagement and move them to healthy users?
- How can you prevent them from churning? Look at your transitioning out users. How can you resurrect them? How can you inspire them to show healthy user behavior?
When answering these three questions, think outside the funnel. Don’t paint yourself into an optimization corner by forgetting about the tornado.
Of course, experimentation is about more than three questions, but these are a good place to start if you’re looking for high-value wins and/or insights.
I read a quote from Dillon Allie of HDMZ recently, which I think sums it all up perfectly…
The buyer’s journey is not really a linear path anymore. It’s more about being ready with the content that prospects need when they are making a decision.
Here’s how you can do that…While linear funnels are easy to understand and communicate, they are simplified versions of reality. Reality is more like a tornado; users don’t take linear paths.
- Tornadoes mean growth attribution is more difficult, but through proper UTM tagging and your Multi-Channel Funnel reports, you can get a detailed view of actual visitor paths.
- Tornadoes mean more points of optimization, which makes prioritization even more difficult. Segmentation goes a long way because you need to optimize actual visitor paths, not intended visitor paths (i.e. the linear funnel).
- Identify user behavior. Conduct conversion research to answer two questions. How do users activate and engage? What is healthy user behavior? A correlative healthy user behavior metric can help answer the latter.
- Configure your analytics setup properly and establish benchmarks. You should know how often visitors are currently activating and engaging if they’re healthy, at risk or transitioning out.
- Conduct experiments to: encourage healthy user behavior, activate and engage at risk users, and prevent users from churning.
Join the conversation
Add your comment
Fantastic article. You found the end of the thread and unravelled the information ball of string so I could see the whole thing more clearly than I have ever done.
Thanks Domenic. Glad you enjoyed the article!
I’ve seen form submissions (which some companies consider a conversion and rightfully so for their business model) increase and conversions (sales) decrease simultaneously time and time again. When optimizing a landing page I can easily target the wrong persona/group and increase form submissions making my new page look like a real winner when it is in fact not attracting high quality traffic.
Failing to target the right people can result in huge declines in your backend conversion. Not all visitors and leads are equal. Utilizing tags is so important. Lumping everyone into groups is a good start, but if your business has anyway to pass revenue back to your original lead or visit do it! If you’re using Ads you should edit your conversion tracking code to pass back revenue dynamically. This can be done by adding variables from your database to the Ads tracking code.
Great read Shanelle. What a line up industry experts here. Nice!
Thanks Evette! Yes, absolutely… I was fortunate to hear back from so many growth experts.
The interest of the linear model is also that it represent what the ideal world would be. A perfect way to understand customer best experience and try to act on your funnel to make it always easier to go through.
If it is not that simple, data and omnichannel experience are required to create some consistency (and make it easier for each segment).
Interesting perspective, Alexis. Thanks for sharing!
I’m starting to enjoy your articles. This one was really good!
Thanks Julien! Awesome to hear.
Fantastic article! Really great information that gets you thinking about all the variables that affect the customer buying process. I can’t wait to throw VR/AR in the middle of all of this to see what the future sales funnels are going to look like..
Thanks for the kind words, Marcelo. Glad you liked the article.
Comments are closed.