You walked away from part 1 of the Growth Hacker’s Playbook with a simple definition of what growth hacking is, a clear profile of what a growth hacker looks like, and when to try to growth hack your business. Now you need some tools to help you actually do it! Your tool belt is empty so let’s start filling it up.
Analytics is at the core of pretty much everything a growth hacker does. The framework for solid growth hacking analytics includes an understanding of people types and the 5 most important growth related concepts for your business: funnels, conversion rates, cohorts, customer acquisition costs, and customer lifetime values. These pieces all fit together like a perfectly harmonized band. If one piece is missing the entire ensemble gets thrown off. Could you imagine listening to Don’t Fear the Reaper by Blue Oyster Cult without the cow bell??!!
To highlight concepts and calculations in this post, I’m going to use examples from two fake triathlon companies I own called “CubedSport” and “TrainCubed”. CubedSport is a triathlon store that sells equipment and clothing and has a physical brick and mortar location as well as an online store. TrainCubed is an online SaaS system with a monthly subscription that gives triathlon training guidance and tracks training results.
People Types Are the Foundation of Analytics
Who uses your website? Unique Visitors, right?
No? You sure? Because that’s what most people use when analyzing the performance of their website, their product or their service.
Let’s pretend for a moment that we don’t live in an online world. When I am working in CubedSport’s retail store in Washington DC, who do you think it is that walks into the store and buys merchandise? Is it a “unique visitor”? Of course not, that sounds ridiculous. I wouldn’t go home at the end of the day and tell my wife there were 149 unique visitors in the store today.
People walk into my store, and those people have a name! People talk to my sales staff, people buy merchandise and people come back to the store at a later date to buy more stuff. Do you see where I’m going with this?
It’s PEOPLE that buy and use the products and services that we sell.
Even though people is a more accurate way of thinking about who uses your product, it’s not quite narrow enough for effective analysis of your business. These “people” come in different forms, shapes and sizes, and to properly analyze them, we need to be able segment them in some way to gather insights from their activity with our products.
People Types provides the segmenting framework that will allow us to do this. People types are comprised of two attributes, an Activity Level and a Frequency. The Activity Level is determined by the type of contact this person has had with your product. The activity levels that I use for TrainCubed (SaaS analytics) are Visitors, Members and Users, and for CubedSport (eCommerce) they are Visitors, Shoppers and Purchasers.
In a SaaS environment, a Visitor is someone who has visited your website but taken no other action with you. A Member is someone who has signed up for your website. Signed up could mean subscribing to your newsletter, registering for a webinar, or signing up for a free trial of your product. Last but surely not least, a User is someone who has signed up for your website and then activated after they signed up. This might mean they attended the webinar they signed up for, or they logged in and used your product after signing up. For TrainCubed, someone is a Member when they sign up for our 14 day free trial, and they become a User after they log their first training session.
These activity types will likely look a little different if you’re talking about eCommerce. A Visitor is still someone who visits your site and takes no other action. A Shopper is someone who visits your site and views a minimum number of products. A Purchaser is someone who has fully completed a cart checkout.
Frequency is the second half of the People Type equation and is the same across both SaaS and eCommerce businesses. The frequencies are New, Returning Free, and Returning Monetized.
New is pretty self-explanatory… it’s someone who has never been to your website before. Returning Free means the person has been to your website before, but has not yet paid your company any money for products and services. An example of a Returning Free person is you if you’ve read my blog before but have not yet paid me any money for the pleasure of doing so J. Returning Monetized is someone who has been to your site before and either currently, or in the past, has paid you money for products or services. An example of Returning Monetized for SaaS is someone who is currently subscribed to your product. For eCommerce, Returning Monetized is someone who has purchased from you in the past and is now back to buy more, hopefully.
A Person Type is simply a combination of an Activity Type and a Frequency. Using the framework we created above, some person types for your company might be New Visitors, Retuning Free Shoppers, Returning Monetized Users. This is a fairly simple concept, but is unfortunately one that is often ignored when doing web analytics.
Just a side note, the definition of what qualifies someone’s activity type as a Member or User will vary based on your company, and that’s ok. All I’ve tried to do here is provide a framework for you to use in thinking through the right terms for your business.
Growth Hacker Weaponry
Thus it is that in war the victorious strategist only seeks battle after the victory has been won, whereas he who is destined to defeat first fights and afterwards looks for victory. ~Sun Tzu
I’m always at war. Growth hacking is a war, a war against customer’s brains, against other companies in my space, against the people that give me budget to grow the product I’m working on. I lose some battles and win some battles. But as Sun Tzu says, my best chance at winning is having a winning strategy prepared before the battle ever begins. Crafting that winning strategy requires weapons, better weapons that my enemy has, but I don’t mean guns and tanks and explosives…
I mean data and statistics. Not fluffy stuff like website pageviews and email open rates, I mean growth hacker weaponry.
There are 5 weapons that every growth hacker should know and completely understand. And by understand I mean understand the concept, and more importantly how to use the number for analysis of their business.
The growth hacker’s weapon armory includes: funnels, conversion rates, cohorts, customer acquisition costs, & customer lifetime values. (click to tweet)
Understand the concept: A funnel is a series of sequential steps within your website or product that lead to a final valuable action for your business. (click to tweet) This final valuable action is commonly referred to as a “Goal”. For example, on CubedSport.com one of my goals is for a purchaser to return to my website and review the product they bought. The image below shows the series of sequential steps someone needs to take to complete my “Review Funnel”. They need to visit my site, go to the main review page for a product, click the review creation button, then submit the review and it needs to save to our database to complete the funnel.
In this example, the review saving to my database is the goal. I could also have created a Thank You page that a person is directed to after they submit a review and that page could have been my goal. Regardless, you always want the goal to reflect full completion of your funnel.
An example of a bad funnel end point goal would have been someone clicking the submit review button. This isn’t a proper goal because there may be instances when the review does not correctly save to my database. Or, in my case, since my site requires someone to log in to leave a review, the person may have filled out the review without logging in, then tried to submit the review but abandoned the process before logging in to complete the review submission.
How to use this for your business: The concept of a funnel forces you to think about your website or product in terms of flows instead of discreet actions. Instead of thinking about a specific discreet action on your site, like clicking the About page, think about how people might use your site to reach a goal that you want. People don’t use website and products in discreet actions, they use them in series of discreet actions that are tied together in a sequence. Recognizing the most important flows and funnels for your site will help you concentrate your analytics efforts on those specific experiences. More on how to use these funnels in a bit.
Understand the concept: A conversion rate is the percentage of people that progress from any one step in your funnel to the next step. (click to tweet) In the image above, 93 people clicked the review creation button in step 3 of my funnel and 76 people successfully submitted the review in step 4. Simple division of 76/93 gives me a conversion rate for step 3 of my funnel of 81.7%. Said differently, of all the people that clicked the submit review button in my Review Funnel, 81.7% converted to actually submitting a review.
You can also have an overall conversion rate for your entire funnel, but depending on where you start your funnel the number may not be very helpful. For example, the funnel in the image above shows that 24,666 people visited my website. If I use visiting my website as the start of my funnel then the 76 people that converted give me a funnel conversion rate of 0.31%. Since I have so much other traffic on my website from people shopping, what might be a more useful metric is knowing that 508 people initiated the review funnel and 76 finished it giving me a conversion rate for that funnel of 15.0%.
How to use this for your business: Conversion rates are your most basic weapon as a growth hacker. Understanding what the conversion rates are at different points in your funnel is your conversion optimization guide. Knowing and managing your conversion rates becomes particularly important if you are doing any paid acquisition marketing (Google AdWords, Facebook ads, etc.). Without understanding conversion rates you may never know if any of the clicks you generating from your ads result in any valuable customer actions.
For CubedSport and TrainCubed I keep a close eye on my most important funnels and the conversion rates at each point in the funnel on a daily basis.
Understand the concept: A cohort is all of the people that moved through a funnel, or completed a conversion during the same time period.(click to tweet) For example, a cohort that I track for TrainCubed is free trial sign ups by month (i.e. all January sign-ups, all February sign-ups, etc.). Each month of sign ups is its own cohort.
A cohort analysis tells you how many people in a cohort, over time, have completed a specific action. Using my free trial sign up cohort as an example, I want to know what percentage of these free trial members are returning to my website by week after the week they sign up. This gives me a high level idea of how useful people generally find the tool. Analytics tools like KISSmetrics allow me to complete cohort analyses easily and present the results in a nicely formatted table. Here is a view of what my cohort report might look like:
This report is showing that in the week of February 22nd there were 34 people that signed up for my free trial. 17.6% of those people returned to the site again within 1 week of sign up, 2.9% returned to the site 2 weeks after sign up and so on. For the week of March 1st, 20.6% of people that signed up returned to the site 1 week after sign up and 11.8% 2 weeks after sign up. That tells me whatever feature update I released on the site the week of March 1st had a large positive impact on my free trial usage. It gives me an indication that I should do further analysis to figure out in more detail what feature was released. Maybe it was something I developed based on some customer research that I did and this usage is validating what I learned from that research.
How to use this for your business: The main benefit of using cohorts is they enable you to track the activity of a defined group of your users over time. Here are some more examples of cohorts I track, and cohort analyses I regularly look at for CubedSport and TrainCubed:
- Cohort: TrainCubed free trial sign ups; Analysis: I look at how often people return to use TrainCubed. This gives me high level view of how good my retention is over time. As I release new product features this basic analysis helps me see if new features are increasing retention and engagement. Let’s say the 5 week returning visitor activity for the cohort of people who signed up the week of April 21st was 90%, 50%, 45%, 30%, 5%. I then release a new product feature the week of April 28th and the activity is 90%, 75%, 70%, 60%, 45% then I know I’ve released a feature that people find really valuable.
- Cohort: TrainCubed free trial sign ups; Analysis: Since I test different free-trial lengths (7, 14, 30 days), I want to know how those trial lengths affect how quickly people enter their first training session into the system. I test each sign-up length for two weeks at a time and create a cohort for each free-trial length. I then create a report that has the cohort on the y-axis and the first training session entry on the x-axis. I can essentially then monitor my products activation curve.
- Cohort: CubedSport eCommerce purchasers; Analysis: Customers don’t come easy and they usually don’t come for free. In one way or another I’m paying to acquire my eCommerce purchasers, so getting them to repeat purchase is extremely important. I create cohorts by week to look at the repurchase behavior of my customers. This kind of report will show me the average percentage of people that make a repeat purchase one, two, three, etc. weeks after their first purchase. Based on other analyses, I know that my best customers are ones who make a repeat purchase within 4 weeks of their original purchase.
- Cohort: Paid TrainCubed sign ups; Analysis: Customer Support is an important part of my business. BUT, customer support does not necessarily mean having people on the phone to answer questions. I’ve invested a lot of time and money in developing and optimizing high quality video tutorials, an online knowledgebase and onboarding drip email campaigns. My goal is to create as scalable a business as possible, and that means minimizing my user’s need to contact customer support. To make sure I’m getting a good return on my customer support investment, I use cohort analyses to monitor how often customers have to contact our support desk over time. A cohort analyses like this enables me to see how groups of customer who sign up in a given time period compare to groups of customers that have signed up in the past.
Customer Acquisition Cost (CAC)
Understand the concept: Your company’s customer acquisition cost (CAC) is the amount of money it costs you to acquire a new user or purchaser (ecommerce). Like I mentioned above, a user can be someone who is either using your product for free or paying you for it. The important part is that they’ve signed up for your product or service and have activated their account, or in the case of ecommerce have purchased a product.
CAC is calculated as the total spend for customer acquisition (ppc, content mtkg, etc.) in a given month, divided by total new customers acquired/activated in the month. (click to tweet)
I’ve used a month timeframe in this definition, but CAC can obviously be calculated on a weekly, daily or even hourly basis. The reason I’ve chosen to define the metric on a monthly basis is to remove abnormalities in data that could occur when calculating it on an hourly, daily or even weekly basis.
How to use this for your business: CAC is part of the customer lifetime value calculation that we’ll go through next. It’s important to note that CAC can be a bit of a funky metric because every company can choose to include or exclude whatever expenses they please in the calculation. For example, for TrainCubed I could calculate my customer acquisition costs as purely the amount of money I’ve spent on CPC ads for the month plus the salary of my in-house blogger. Or, I could also include the costs of Unbounce (which I use for landing pages), and part of my salary because I spend a portion of my time managing the marketing campaigns, and a portion of our server costs because our marketing efforts drive traffic to our site and we have elastic cloud servers set up via Amazon. My point here is that CAC is a valuable metric as long as you understand what is included in it and keep that calculation consistent over time for comparative analyses.
Customer Lifetime Value (CLV)
Understand the concept: Your company’s customer lifetime value (CLV) is the total monetary PROFIT that an average customer represents for your business. This is one of the top 5 most important metrics for your company, if not the most important.
CLV is calculated as the total revenue a customer generates for your company (the sum of revenue you’ve received from this customer for every month from the time they signed up to the time they churned or made their last purchase), less your CAC. (click to tweet)
This concept sounds simple in theory and I’ve presented it that way intentionally because I want everyone to be able to understand it and use it to at least some extent in the operation of their business. However, this simple definition makes two very big, rather unrealistic, assumptions. The first assumption is that once you acquire a customer there are no customer retention costs. The second is that the customer doesn’t at some point leave you and then come back as a re-acquisition.
I’d venture that no business operates in an environment where customer retention costs are $0. (click to tweet) So I’ve left you between a rock and a hard place, up the creek without a paddle, or any other cheesy metaphor you can think of. Now what…?
If you are new to analytics and customer valuation, then I suggest you use the Revenue – CAC version of the equation and know that you are doing a lot better than most others.
For the more experienced growth hacker, marketer, startup founder, CEO, and business analyst, you should look at your P&L and company’s operations and make an estimate of your customer retention costs. Be sure to include allocations for your things like your email platform, customer support reps, the customer support technology platform, your knowledge base upkeep, designers, copy writers, and paid re-engagement marketing to name a few.
Those of you who are more advanced can take Revenue one step further as well and include allocations for referral program revenue and display ad revenue if applicable to your business.
How to use this for your business: At the risk of sounding like a broken record, CLV is one of the top 5 most important metrics for your company, if not the most important.
Any growth effort should be evaluated by how it will affect CLV. You want to run an acquisition marketing campaign? You better be able to acquire a new customer in a way that results in a positive CLV. You want to develop a new feature website? The costs to do so should be covered by additional revenue generated from existing customers (therefore increasing their CLV), or from new CLV positive customers that will be acquired as a result of it.
Take a deep breath and hold onto your socks, the discussion of CLV isn’t over quite yet. One last thing to keep in mind is that the timing of the revenue generated from a customer is very important. For example, your customers may have a lifetime value of $200, but it costs you so much to acquire and retain them that from a cash flow perspective you are losing money on them for the first 2 years.
If you are a small startup or a company with limited funds this is crucial. It’s crucial. One more time, it’s crucial that you understand what this means. It means that if you aren’t generating revenue from your customers quickly, you could bankrupt your business despite have positive CLVs.
Here is an example of two companies with comparable products, the same CAC and positive CLVs, but very different cash flow profiles:
For this example let’s assume that the customer churns after month 12 for both companies. Company #1 offers a $15 monthly payment option, and while it has a CLV of $56, it takes 8 months for the customer value to be positive. Company #2 offers a $150 annual plan, and while at $26 the CLV may be lower than Company #1, Company #2’s customers have a positive cash flow profile starting on day 1. What a great advantage to have in the market for new customer acquisition and overall financial health of the company!
Here are some examples of how I use CLV to manage CubedSport & TrainCubed:
- The current lifetime revenue for my average customer is $300. My customer acquisition costs are $100, and my retention costs are $150, resulting in a $50 CLV. When I think about running a new paid marketing campaign, I know I only have $50 to play with in my CAC. If my estimate for CAC for this new campaign is $160 I’d end up with a CLV of -$10, that’s bad. But, if I think these newly acquired customers will have higher lifetime revenue than my current customers, and will more than cover the cost of their higher CAC, then I could be in good shape. In this scenario I’d do a test of the campaign and look at the first few months of revenue generation so I can estimate the long-term effect on my customer lifetime revenue and thus the CLV.
- I was recently evaluating a new customer referral program for CubedSport. Without knowing my CLV I may have just copied others in my market and offered $20 to the referrer and $20 to the referee once the referee made their first ecommerce purchase. But, since I knew that my CLV was $50, I took a chance and structured the program to give $20 to the referrer and $30 to the referee. This puts me in a better position relative to my competitors while not allowing my CLV to dip negative, and gives me the opportunity to test whether these newly acquired customers will have higher lifetime revenue because they were referred by a friend.
Bringing It All Together
Well, my intention was to get through the significance of people types, each of the pieces of growth hacker weaponry, and a guide on setting up analytics for your company all in one post. But, since I’m quickly approaching 4,000 words, I get the feeling you may have had enough for the moment.
Analytics is not easy, frankly neither is growth hacking. If they were, everyone would be doing both really well. Using People Types and your set of growth hacker weaponry (funnels, conversion rates, cohorts, customer acquisition costs, and customer lifetime values) will help you analyze what people are doing on your website and with your product, and will give you the ability to make smarter decisions for your business.
Please know that even as an experienced growth hacker, analyzing data can be an overwhelming and even crippling exercise. If you’re new to this, start trying and it will get easier over time. Concentrate on one area of your business, one funnel, one cohort or just one conversion rate at a time and it will reduce your feeling of being overwhelmed. If you have questions please reach out to me via comments below or by email at email@example.com.
In the next post I’ll walk through setting up analytics for your company including a review of current platforms (broad, person based, marketing specific, and email specific) setting goals for your product/service, A/B testing and the difference between user centered design and conversion centered design.
Do you already have an analytics framework set up for your company? How does it differ from mine?
Cover image credit: Nexus Wallpaper