Using Segments to Optimize Customer Value: Advice from Keith Perhac

Not all customers are created equal. Because of this, you need to identify which ones to prioritize. But how exactly do you do that in practice? According to Keith Perhac of SegMetrics, it's about using segmentation:

"What you want to do is figure out which segments are good, which segments convert and which ones don't, and then get your customer value out of that."

Keith shares how to segment your customer base, evaluate your marketing efforts using segmentation, identify the customer actions that influence key SaaS metrics, and more.

Watch the recording below for the full video version, or read on for the bite-sized learnings.

Two key dimensions: segments and touchpoints

Most SaaS metrics platforms segment users by product or plan. While it's easy to understand, it doesn't really give you any unique insights or learnings about your customers. This is why Keith and his team at SegMetrics emphasize on these two important aspects of users:

• Customer segments
• Unique touchpoints on their journey

Let's talk about them in detail.

Unique customer segments that exist for your audience

While there are many ways to segment your customers (ie. interests, type of business, level of sophistication), Keith shares that the main ones they look at are industry and acquisition channels.

Unique touchpoints on your customer journey

You also need to look at the things users do on your customer journey. This will pinpoint you to what could be improved in your process.

For example, SegMetrics customers go through the following touchpoints:

• Acquisition channel (e.g. Facebook ads)
• Trial
• Setup
• Kickstart call
• Becoming a paying customer

When looking at these two aspects, the goal is to find the best combination of touchpoints and segments.

How to narrow down on the best combination

Step 1. Define your baseline metrics

The first step is to look at the baseline metric. In this example, we see that there's a 54% conversion rate from trial to customer. But what else is there to learn?

Keith says this is exactly the problem with these baseline metrics and that's why you need to break it down further if you want to know the problem:

"This is the problem with all top-line KPIs. It's like a speedometer: you know what's going on but you don't know why it's happening."

Step 2. Separate each touchpoint through the journey

The second step is to look at the different touchpoints in your journey, and your metrics on that level. This is where you might see the similarities and differences in your touchpoints.

Step 3. Identify outliers

From these data, you then identify your outliers.

In the example above, we see that not getting on a kickstart call is a big blocker in the road to conversion. In contrast, people who are getting on the kickstart have a higher chance of becoming a customer.

The next question would be: why are people not getting on that kickstart call?

Step 4. Identify performance at a segment level

To answer the previous question, you now have to identify the performance at a segment level. Which segments are having a problem with that touchpoint?

In this example, we see that agencies are converting very well, while users from ecommerce aren't.

Now you've identified the unique segment that's having trouble with the kickstart calls, and you can start thinking of solutions to help them move forward. It can be offering them a $25 gift certificate, a better self-serve option, or have an entirely different flow for this segment.

Keith emphasizes that outliers are the levers you can pull to optimize your marketing:

"We have to find the outliers. And that's why we found ecommerce and agencies here because the outliers are the things that are going to let us optimize our marketing and target the users that are doing extremely well or extremely poorly and we can make a decision there.

The people who are doing well: cool, we do more of that. The people are doing poorly: we have to make this decision. Are they people that we want in the app? Are they good customers or are they bad customers? If they are good customers, but we're not serving them well, how do we change the offering in order to serve them?"

When it comes to cohorts, it's best to get specific with them, so you can find that valuable audience:

"Broad cohorts obscure both good and poor performers. If we looked at that 35% and 32%, there's no difference between the two. We don't know what's going on. If we looked at that top-level 54% conversion rate, we don't know what's good or bad because it's a broad average. We have to get specific with our cohorts to find that valuable audience."

Step 5. Look at the lifetime value by segment

While conversions are a great way to start, it's not the end goal. This is why you have to take your analysis a step further. You have to look at the lifetime value so you can identify the segment that you need to focus on.

In the example, we see that customers from the info products industry are worth more even though they are only at a 35% conversion rate. This means that even though infoproducts and SaaS are at the same conversion rate, customers from the infoproducts stay on much longer and are much more valuable.

Optimizing for valuable customers

To do this, we need to optimize the lead value of each segment and touchpoint. The lead value is the average amount of money generated by anyone on your list at a specific segment or touchpoint.

To calculate for the lead value, divide the amount of money earned by the number of leads in that segment:

So let's say you have a hundred people in that segment and you made a $100 in revenue. Each lead is worth $1.

"You increase your lead value by selling more to the same number of leads or by increasing the number of people who purchase. You increase your conversion rate or you increase the lifetime value by how long they're on your subscription product, how many products they buy from you, buying the bigger product, things like that."

Keith even goes a step further by getting the touchpoint lead value, which puts a dollar value on each touchpoint in your funnel.

"It takes our conversion rate and gives it an actual dollar value. And that's what we want to optimize towards not how many people did this, but how much was it worth to do that?"

In this example, we see that Facebook ads give better returns than another acquisition channel — which means that you can scale this channel effectively. Because there's a 126% return on the ad spend, this also means that this is now a growth factor.

Additional marketing tips

Tip 1. There will always be a market size ceiling for ads

While it might seem that you could always get good ROI on paid ads, Keith shares WP Engine's experience which shows that you'll only get it up until a certain point:

"Jason Cohen of WP engine was giving a talk about how they had optimized their ads so well that they knew for every dollar they put in, they were going to get two out. And people were like, 'Well man, you could make an infinite money machine. You can just scale this to the moon.'

And he's like, 'Yes, except we get diminishing returns because support cannot keep up with the new customer.' So they had to limit their acquisition. Because of something further down the funnel and they weren't able to see that just at the ad level."

With the framework that we just discussed, we see that WP Engine had trouble with the conversions and users staying on in that lifetime value because support wasn't able to keep up after a certain amount of new trials and customers. To solve this problem, WP Engine then optimized their ads not only for ad performance and conversion, but for that entire experience that ties into the lead value.

Jane shares that this is hitting the market size ceiling, where very often you can pump up your ads to a certain point, and then it just stops hitting the right numbers.

Tip 2. Increase ad spend gradually

To avoid hitting that ceiling, Keith advises that you have to increase your ad spend gradually because what works with small numbers might not work with big ones.

"If you see an ad campaign, don't just say, 'Oh, it's doing great, let's throw $10,000 at it.' It has to be gradual because otherwise, you're going to hit that ceiling. And you're not sure where that ceiling is at first and the algorithm that is powering it. These ads are used to these small numbers.

So if you suddenly say: 'Oh, here's $10,000', the targeting becomes wider. And so you're not actually targeting the same people you thought you were when you had this smaller audience. So always do things slowly and grow them out or organically, or not all at once because that's a recipe for spending a lot of money and not getting good returns."

Tip 3. Make sure the landing page reflects the same offer in the FB ads

“Look at the main offer in Facebook ads and make sure that the landing page offers the same.”

Tip 4. Ask the right questions to improve your marketing

Instead of searching for best practices, Keith says it’s better to ask the right questions so that the solutions are customized for your business:

“Being able to really look at and discover, ask the right questions about your business. I think this is really the core part of how you improve marketing.”

Tip 5. The duration and cadence of the experiment depend on the length of the campaign

Keith advises determining the duration of your experiments based on the particularities of your journey (ie. trial to conversion) so that you can see the short-term and long-term trends.

“For example, we have a 14-day free trial. We can't look at 14 days that doesn't really give us much information. So we generally look at 18 days for initial conversion, and then we have a number set up for three, six, and nine months. And we have built a dashboard that essentially takes the values in that first trial period. And then the values three months, six months, nine months later for that same period.”

Tip 6. Don’t obsess over the marketing dashboard

While it’s nice to monitor the trends of your KPIs, Keith says it’s better to determine the regularity of reviewing data based on the role you play in the company.

“If you are the sole person in charge of that, and you are doing a good amount of traffic, I would honestly look at it daily. I would have it as your homepage and look at it daily. If you have team members and they are in charge of different aspects of it, I think you should look at it twice a week.

I really think the shortest timeframe that you should look at unless you are in a high transaction business is weekly. There's no value in looking at any segment of time, shorter than a week.”

Tip 7. Track the pre-conversion journey to help you attribute the conversions

For most SaaS businesses, leads that convert usually come from Google. This gives us an idea that the long-tail marketing probably works. But how do you attribute these leads?

Here’s what you can discover when you track the pre-conversion phase of the customer journey:

“Even before they sign up, they'll type ‘userlist’ in the search bar. They'll click on that Userlist ad because it's the branded search. And then they convert because they've been thinking about it forever, but you don't know what pages they were on.

We actually track all that information beforehand. Then when we see that person convert, and we turn that all into that singular customer journey. So we have data from like three months before they ever converted.”

Tip 8. For SaaS businesses, it’s better to invest in Google ads over Facebook ads

Drawing from his experience, Keith recommends Google ads over Facebook ads for SaaS businesses because the latter is not the best channel for conversion. This is because when people are on Facebook, they are looking to be entertained instead of looking for solutions. In contrast, most users go on Google to look for answers to their problems.

Based on their attribution data, they found that Facebook ads can still point customers in the right direction during the pre-conversion stage of the journey:

“Facebook is essentially a billboard for us. No one is signing up from Facebook, but they're all seeing it.”

Tip 9. Make your Google ad copy solution-based

Because most users are looking to Google for answers to their problem, Keith says that you have to make your ads solution-based to get good conversion. And the best way to do this is to make it as simple as possible by answering “the 2:00 am problem.” For example:

“Let's say you're in the health space. You're like: ‘Why do I have diocardiatha (not a real medical illness)? No one's searching for that. At 2:00 am, what is the pain point? What is the lizard brain pain point that they are searching for? ‘Why does my stomach hurt so much?’ — that’s your ad.”

Tip 10. Make “alternative to” pages and run ads for them

A super easy tip to make good converting Google ads and ranking pages is to create “alternative to” pages for each one of your competitors.

“Write a page for all of them. ‘Alternative to X’, ‘alternative to Infusionsoft’, ‘alternative to Activecampaign’, ‘alternative to MailChimp.’ Run ads for them.”

Learn more from SegMetrics

SegMetrics is an attribution and business intelligence platform to help you get clarity on where your leads come from, how they act, and how much your marketing is really worth. If you want to learn more about how to optimize your marketing efforts, you can look at their free marketing courses.