Kissmetrics Blog

A blog about analytics, marketing and testing

Built to optimize growth. Track, analyze and engage to get more customers.

Marketers: 5 Ways to Kick Your Guesswork Marketing to the Curb with Analytics

Let’s be honest, most of us have no idea how well our marketing does. We choose a target market with some random demographics, pick a channel, and throw a campaign out into the world hoping for the best. Even after the campaign has run its course, we have to keep hoping. Why? Because we usually don’t have a way to figure out if it worked even after it’s finished.

More often than we want to admit, our marketing budget runs on hopes and dreams.

But it doesn’t have to be this way. With just a pinch of effort, we can add accountability to our marketing.

I’m going to show you how customer analytics gives you:

  1. ROI on all your marketing campaigns and traffic sources
  2. Lifetime value of all your customers
  3. Ability to track branding campaigns
  4. Help you find the right customers
  5. Make sure you sell the right benefits

Yup, it does all that. So let’s dive in and show how each of these works.

Return on Investment

The entire point of marketing is to bring in more customers. That’s it. Branding, direct response, and PR all bring you more customers either today or tomorrow. At least, they’re supposed to.

And those customers have to bring you enough profit to cover the cost of the marketing. If they don’t, you’re losing money and you’re not growing your business. In this case, you should just fly your whole team to Europe for the quarter. You’ll burn the company coffers just as fast but you’ll have a lot more fun.

How do we know if we’re on the right track? ROI (return on investment) keeps us in check.

Basically, you figure out the total revenue that you received from your marketing. Then you subtract cost of goods sold (what it costs you to get the product or service), which gives you the gross profit from the campaign. Now is this number bigger or smaller than what you spent on the marketing?

If your marketing isn’t pulling in enough revenue to cover your cost of goods sold and the marketing budget, you have a negative ROI and the campaign was a failure.

So the first step is to figure out how much revenue your marketing campaigns bring you.

And guess what? This is built right into KISSmetrics. Have a look:

KM Campaign Revenue

As a marketer, I LOVE data like this. So here we go, we’ve got:

  • Total revenue from the entire campaign
  • Average revenue per person thus far
  • The lifetime value that I can expect from each person
  • Total number of customers I’ve received from the campaign
  • The churn rate

This is better than Christmas. In fact, I don’t want any presents this year. Just keep feeding my reports like this one and I’ll be one happy panda.

To keep things simple, start with that total revenue:

KM Campaign Total Revenue

Right away, we know the total dollar value that the “cohort upsell” campaign has given us. Down to the last dollar. To figure out whether or not the campaign made us money, there’s another step. I don’t want to scare you but we’ll need to do a little math now. Don’t run away! I promise it’ll be easy!

Let’s say you run a teddy bear company and you have an ecommerce store. And this is your data. You’ll need to go to the finance wizard on your team and find out how much it cost you to get all those teddy bears (cost of goods sold). Let’s say it’s $13,000 worth of teddy bears.

Now we need to find out how much we spent on the cohort upsell campaign. It was a mix between AdWords, Facebook, and some banner ads. The total cost was $4,000.

So you spent $17,000 ($13,000 + $4,000) to make $30,078. Not bad, not bad at all.

Now you know that you have a profitable marketing campaign. High fives!

Bonus Round: ROI calculations can get MUCH more complicated than this. I’ve also simplified the finance side of things. If you want to jump into the deep end of the finance pool, go for it. But when you’re just starting, this will tell you whether or not your campaigns are working.

You’ll want to be careful when pulling revenue data from most web analytics products. They were built to track your website (instead of your customers), so there’s no way to piece together multiple purchases.

In other words, someone finds you with a Google search and that keyword gets the credit for the purchase. But then when they visit a second time from your email list and purchase, the credit goes to your email campaign. One customer, two purchases, and credit is split between two different traffic sources. When you’re using customer analytics, this isn’t a problem at all. Every purchase gets tied to the same customer, so it’s easy to see the total revenue that traffic sources are bringing you.

Ideally, you want to be able to sort your traffic sources and your campaigns (along with the revenue they generate) by the first touch point. It’s the first campaign or traffic source that brought the customer to you. This way, you’ll have the total revenue generated from traffic sources and campaigns. Even if they make multiple purchases.

Are the reports above sorted by the first touch point? You betcha.

I can also sort them by the last touch point to see which of my campaigns close the sale.

Just about every decision you make as a marketer ties back to ROI. Are we spending too much time on Facebook? How much can we spend to acquire a customer in AdWords? Did the $10,000 we spent on SEO consulting pay off?

Get a handle on your ROI and all of your decisions get a lot easier.

Lifetime Value

Lifetime value (LTV) is the holy grail of marketing analytics. No joke. This is possibly the most valuable metric you’ll ever use.

Before we jump into using LTV, we need to get our heads around some of the other metrics in this report:

Average Revenue Per Person: This is the average amount of revenue that you’ve already received from that group of people to date. They’ve already paid that amount and the money hit your bank account.

Churn: The percentage of customers that stop paying you in a given time period. Our reports use a monthly churn rate.

Lifetime Value: This takes the average revenue per person and looks at the churn rate. Based on the revenue coming in and the rate that customers are leaving, it predicts the total revenue (lifetime value) that you can expect to see in the future from that group of customers. While average revenue per person tells you what you’ve already earned, lifetime value tells you what you can expect to earn.

Let’s pull that revenue report again (I’ve highlighted the LTV action):

KM Campaign Revenue LTV

Well that’s cool, the “cohort upsell” campaign and the “cohorthd” campaign have a very similar lifetime value of about $7,500. So if we’re only trying to maximize for long-term revenue, it really doesn’t matter which campaign we keep using. We’ll end up in the same spot.

But let’s dive a little deeper. Here’s the average revenue per person and the churn rate for each:

  • Cohort upsell: $198 average revenue per person and 0.2% churn rate
  • Cohorthd: $588 average revenue per person and 0.8% churn rate

This means that the cohort upsell customers aren’t spending as much money but more of them are sticking around. On the other side, we have the big spenders of the cohorthd that leave faster.

Here’s the thing: if we lowered the churn rate from the cohorthd customers by just a touch, it would shoot the LTV through the roof. We have a decent number of customers that are spending a fair amount of cash, but aren’t feeling as valued as customers from our other campaign. Let’s reach out to the ones that already left and figure out what we could have done better. With a few tweaks and fixes, the cohorthd campaign could quickly become our most profitable campaign.

Even a simple thank you phone call to those 51 people might do the trick.

By focusing on the LTV of different customer campaigns, we know exactly what to work on in order to keep growing the business.


What’s the fastest way to get your marketing team to mutiny? Ask them to measure their branding.

Let’s cover how branding is typically measured.

You’re about to launch some snazzy branding campaign nationwide for a Fortune 500 client. And you’ve had the foresight to test the campaign in one city to see how well it does before going after the entire US pie. There’s two ways marketing teams tend to measure the results.

  1. Total impressions: simply get as many eyeballs on your campaign as possible. You’ll use metrics like pageviews, TV viewers, and newspaper subscribers to get an idea on how many people saw your campaign. Then you’ll calculate the cost of impression for each and try to get that as low as possible. A low cost per impression equals success.
  2. Ability to recall: using surveys, you’ll ask people in your target market whether or not they can remember your brand and the campaigns that went with it. No recall is bad, lots of recall is good.

And yes, there’s a major flaw with both these approaches. Neither one ties the campaign back to an action. Great, people saw the campaign and they remember it. So what? Are they motivated to reach out to you? Will they come to you later when they have a need for your product?

I’m going to let you in on a little secret: there’s a better way to measure branding.

You see, as soon as you start using customer analytics ( *cough* KISSmetrics *cough*), you can track people through the ENTIRE lifecycle. So if someone engages with a branding campaign this month and it takes them another 17 months to become a customer, you can track that.

You can even track how good a job you’re doing at building a relationship with them. Take a look at this cohort report (click on the image below to see a larger view):

KISSmetrics Cohort Analysis

Before we jump into what this data tells us, let’s cover what it all means.

On the left, we have people that have been tagged by a campaign. And they’re broken up by month so that we can compare different batches of people to each other.

On the top, it lists the number of weeks that it took for someone to visit our site.

And the percentages are cumulative. So all the people that visit our site 2 weeks after being tagged with a campaign also get added to the total from week 1. This way, we can look at the last column (>12 weeks) to see the total number of people that came back from a campaign.

When we have data like this, it’s very easy to see which months had branding campaigns that helped build awareness with our target market. In this case, the seasonal campaigns we ran in October and November of 2011 did a great job at building awareness. Over 85% of those people came back to us.

Now branding is easy to track.

What about looking at individual campaigns? Yep, you can do this right in the cohort report (click on the image below to see a larger view):

Cohort Campaign Name

On the left, we have all the campaigns that sent people to us. And the cohort report is tracking how many people from those campaigns came back over time.

With a quick glance, we know that the sendgrid campaign is REALLY driving engagement. People are coming back day after day, especially compared to our other campaigns. Let’s look at it in more detail:

Cohort Sendgrid Campaign

So within a single day, 16 of the 27 people from the sendgrid campaign came back. Looks like we struck a nerve with that campaign.

Now we know which branding campaign is giving us the best engagement. Granted, these are small numbers (from a real company). But even if you’re moving at a much bigger scale, think of what you could do with data like this. You’ll know how well every campaign works, even if it doesn’t have a defined sales goal.

Ready for the best part: you can structure these cohort reports however you want. Any data, any campaign, any length of time. You have complete control over your data. Want to track how many customers (not just visitors) you get from branding campaigns? Done. Want to see whether or not social media is worth the effort? Not a problem. It’s your data and you should have complete control of it. With KISSmetrics, you do.

Where to Find The Best Customers

Great, we know whether or not all our time on Facebook is actually productive. And it gets even better.

Once you know the LTV of customers from different campaigns and traffic sources, you can figure out what gives you the BEST results.

You see, not all customers are equal. Some of them are a perfect match for your offer. They sign up quickly and stay with you forever. Others suck profit straight from you as your sales team spends countless hours trying to get them to sign up and then the support team is up past midnight trying to keep them happy. Even when you change the earth’s rotation to match their needs, they still bail.

So where are the awesome customers? You already have the LTV of your campaigns. That’ll tell you.

When you’re looking at your reports and you have a list of traffic sources or campaigns along with the LTV for each, you’ll know exactly where to spend your time. Take the traffic source with the highest LTV and go do more of that.

Bonus Round Deux: LTV doesn’t factor in the cost to support the customers. So check with sales and support to make sure that the high LTV customers aren’t sending you endless requests.

And as you experiment with other traffic sources, keep comparing the LTV with your best performers. Once you have a couple of dozen customers from that traffic source, you’ll know how much potential there is and whether or not you should keep going in that direction.

Selling the Right Benefits

Whenever we’re putting together the next promotion PR campaign, we’re constantly facing choices about which benefits we should talk about. Many channels severely limit how much we can say. Even when we have room to go deep, we only have a few seconds to catch the attention of our audience. Sometimes, we only have a single sentence to get our point across.

When every word counts, we need to make sure we’re pushing the right benefits.

But products usually provide multiple benefits. They save time, increase revenue, make customers happy, and bake them a cake. What should we focus our messaging on?

The easiest way to figure what to sell is by seeing which features your current customers use the most.

Here’s a timeline of every action someone has taken with a product (click on the image below to see a larger view):

KISSmetrics Timeline

Yup, that’s right. It includes every last event (listed on the left) that they completed. Each vertical line represents a different day and as people complete the same action repeatedly, the dot for that action gets bigger. So when you track different features of your product, you can see how frequently your customers use them.

In no time at all, we can quickly see which features are most important to this customer. And when we know which features matter the most, we know what benefits to sell.

Want to Get Your Customers Into These Reports?

We’d love to show you how KISSmetrics helps you understand your customers so you can grow your business even faster.

Bottom Line

When you start using customer analytics, many of the typical problems are no long problems. We can go straight to our data, see what’s working, and spend time growing the business.

There’s no more guesswork involved.

So if you want to:

  • Get the ROI on all your campaigns
  • Know the lifetime value of your customers
  • Finally have a way to track branding campaigns
  • Focus on the right customers for your business
  • Sell the benefits that bring the most customers to you

Go ahead and get your hands on customer analytics.

About the Author: Lars Lofgren is the KISSmetrics Marketing Analyst and has his Google Analytics Individual Qualification (he’s certified). Learn how to grow your business at his marketing blog or follow him on Twitter @larslofgren.

  1. Martin Voorzanger Jul 09, 2012 at 10:16 pm

    Hi, once again a great post. There’s one part I don’t get – if I run a campaign on different media/channels (for instance AtL such as TV and radio), how do I know which medium works best. Do you use different URLs for that, or ?

    • I would say yes, utilise multiple landing pages that reiterate the messages and CTA conveyed in your campaign.

    • Yeah, the easiest way to track offline campaigns is with a simple and easy-to-use URL that’s unique to the campaign. You can either redirect that URL into a URL that has all the utm parameters you need or you can set up events on the landing page to track those people. By setting up a unique URL and landing page for each campaign/channel, you’ll be able to see the effectiveness for each.

      A lot of ecommerce site will also use unique discount or promo codes for this.

  2. Great post man!
    Excellent article keep on posting such an informative article with us.

  3. Lifetime value and branding is one of the most important thing in marketing.

  4. Social media marketing also plays an important role in protecting brand names: 58 percent of social media marketing users say they write product critiques to protect others coming from bad experiences, and practically 1 in 4 points out they share their unfavorable experiences to “punish companies”.


Please use your real name and a corresponding social media profile when commenting. Otherwise, your comment may be deleted.

← Previous ArticleNext Article →