How to Measure Your Way to Startup Success

What follows is a guest post by Josh Ledgard.  He’s a KISSmetrics customer and the founder of KickoffLabs.

Every business is based on a set of assumptions. If you were going to open a bagel shop you’re initial set of assumptions might be:

  • You’ve discovered a great bagel recipe that tastes great and is affordable to mass produce.
  • The location you’ve chosen is convenient enough for a sizable target market.
  • You have a marketing campaign that can dislodge people from their current habits that include stopping at McDonalds, making their own breakfast, etc.
  • Once someone has tried your bagels, they have a yearning to come back… and tell their friends to stop by.
  • Customers are then going to stop by at least once a week for the next year and buy at least $5.00 worth of food each visit. :)

There isn’t much difference between these assumptions and those generated by someone creating an online business.  What’s good for you is that online businesses are easier to measure.

comparing online business testing to a bagel shop

In this post, I’d like to share with you a set of best practices we employ at KickoffLabs to constantly measure performance against our various assumptions and theories. We’re a software as a service business that generates leads with stunning landing pages designed to convert campaign traffic into signups, and signups into shares amongst friends. Customers setup a landing page with our tools, promote it, and capture leads they can sell or market to.

Our business success depends on a lot of assumptions about our target market AND about how well our tools will work for our customers. Because we employ a freemium model we also assume that a certain percentage of those customers will upgrade and that the rest will help us get new customers because of the branding we put on free customer pages.  My hope is that by pulling back the curtain a bit and laying out a simple framework, you’ll be able to put together an approach that suits your business, personality, and work habits.

We work constantly to test assumptions that aren’t very different than the examples I used for the bagel shop. Just like any brick and mortar business, we focus on metrics in order to improve upon our success rate.  The name of the game is progressive, incremental improvements.  We believe that you can’t control what you don’t measure. So here are five principles we use to guide us:

1. Generate Weekly and Monthly Reports of Key Metrics

There are so many numbers you could analyze for your business that it’s important you pick the right ones to focus on.  If you’re driving the wrong numbers forward, it’s distracting at best and potentially catastrophic. We’re a subscription based SaaS service and these are the numbers we look at each week:

Our Success

  • Monthly Revenue Projections. Because we’re a boot-strapped business, we need to keep a close watch on cashflow.  Each week we look at current subscription income collected, to be collected, and assumed new customers.
  • Growth in Number of Paying Subscribers. This tells us how well our overall customer acquisition strategies are doing.
  • Overall Conversion Rates. For this metric, we’ve setup a basic funnel from “Viewed” to “Signed Up” to “Paid Conversion”.
  • Top Sources/Referrals that Drove People into the Top of the Funnel. Some we expect to see based on known campaigns and some may be a surprise. (More on this below)
  • New Cancellations. This helps us calculate metrics like the total lifetime value of a customer.  If 10% of our customers cancel each month then we would expect each customer to be worth 10 months times our average subscription price.
  • Results of Marketing Campaigns in Terms of their Specific Conversion Funnels.  Is it worth spending money on the Google Adwords campaigns we invest in?
  • Results of any A/B Tests Run that Week on our Signup Pages in Terms of Conversions.  So we can pick a winner and try to beat that next week. If you aren’t testing you’re going to be failing… but this is a topic for an entirely new post.

Notice that this doesn’t include any measure on customer demographics, bounce rate, time on site, etc.  Those just wouldn’t help us test our assumptions. We could, of course, measure any number of things, but we’ve settled on the factors above as providing the most useful bird’s eye view of how our business is performing.

Customer Success

We believe that no matter how good our funnel is, we’re only successful if our customers are able to run successful campaigns.  So we also have a set of key customer success indicators we analyze including:

  • Customer landing page conversion rate.  We track the conversion rates on our customers landing pages. If our pages weren’t converting well, then we know customers would drop off sooner.
  • Most viewed and highest converting customer landing pages. From this we get to study what’s working well for these customers so we can adapt those best practices for our entire customer base.
  • Customer Shares.  Are people happy enough with our service that they are telling their friends?  If we drive up the percentage of our customers that tell their friends about us, then we know they felt successful regardless of the conversion rate they saw. Seeing WHEN they do this also helps tell us how many conversions most customers view as a successful number. For this we measure when someone shares online from our app using KISSmetrics. This is the most useful, but we also measure Google Alerts to find when a customer blogs about us, and we conduct twitter searches to see when our brand is mentioned positively.

KISSmetrics Funnel Report

Get your key metrics into a funnel report.

2. Don’t Automate Every Report

We use KISSmetrics to automate several of the above-mentioned metrics in order to save us time. However, I firmly believe that manually digging through the data on a regular basis can lead to unexpected insights you’d miss if you just waited for the report. Here are three things we do that are not fully automated on purpose.

I. Drill down into the customer sources. Customer sources would include search engines, web site referrals, pay-per-click ads, and campaigns you execute to drive traffic back to your site. For example, let’s assume you answered a question on Quora and it resulted in a customer upgrade. Because it was just one customer it’s unlikely to appear on the higher-level automated report. Yet identifying that conversion alerts you to an opportunity that might be worth pursuing. Clearly there is something of value in your answer, so perhaps you could run a test on whether more upgrades would occur from promoting that question and answer more.

II. Personally email “extreme” customers at key events.  When someone leaves our service after having either a LOT or no success I’d like to know what worked well and what didn’t to learn how we could improve.

III. “Watch” users flow through your site live. We also use Google Analytics for the real time reporting.  When we know that we’re having a traffic spike, it’s useful to watch customers flow through our conversion and site launch process.  Seeing how the users click around our site live gives us insight like watching from behind a one way mirror.  Yes, you could wait for a flow report at the end of the week… but seeing it live gives you a sense of how long it takes, where they get stuck, etc.

3. Track Where Your Customers Come From

Working out the “origin story” of your customers can reveal important insights into that market. Before getting too carried away, though, it’s essential to point out that you need to concentrate primarily where PAYING customers come from.

Using the data we collect with KISSmetrics we are able to build on report that looks something like this:

Table 1: Trending Customer Sources in a week

Source

New Upgrades

New Upgrade %

All Time Upgrades

All Time Upgrade %

Free Customer Branding

2

12%

20

20%

Customer Referral Program

2

12%

20

20%

Pay-Per-Click Advertising

2

12%

30

30%

KickoffLabs Blog(click to view by posts)

2

12%

20

20%

Quora(click to view by posts)

8

50%

10

10%

KISSmetrics enables our stat-geek Nirvana. :)

When you gather this intelligence, you learn:

  • What’s trending.  Notice the highlighted cell tells me that we must have done something right on Quora.
  • Where you can double down some investments.  In this case, maybe we should answer more questions on Quora.
  • What you may have to scale back on. If you know that you spent a lot of time on your customer referral program, but it’s not generating results… maybe it’s time to cut that.

4. Real Time or Daily Email Alerts

At KickoffLabs we generate email alerts to ourselves for things like new signups, upgrades, new customer sites, etc. Each of these emails contains a snapshot of how people signed up, what pages are working best, and what kind of sites users are putting up. We put that data right into the emails.

In relation to that last point… I don’t have time to read every site that comes through, but I don’t need to; just scanning the email alerts shows me some interesting landing pages that I’d miss out on if I just read a report of the top sites.

account deleted

From these emails, for example, I get a snapshot of the sites the user had created and how successful they were from the number of conversions they generated.

What this process does, over time, is help you internalize a set of patterns for how things have been performing versus how they are doing now. This leads to informed opinions about your business that can be tested against empirical data. It’s subtle, but after a while you end up with your finger on the pulse of your business.

The “Second Conversion” – Referrals

Our funnel doesn’t end at the upgrade.  We need to know how many of our customers are:

a. Sharing Kickofflabs with their friends.

b. Generating new paying customers… so we can reward them.  We generated special campaign links with the Google campaign URL builder to see both how well this program works with KISSmetrics and which customers are generating the referrals… since we set the source as a specific customer ID.  Every time a customer refers someone, we give them a free month of service.

In addition to using KISSmetrics, we’ve also used KISSinsights to measure what % of people would recommend our product or why they wouldn’t.

kickoff labs referral program

We encourage sharing for paid customer conversions.

5. Don’t Get Addicted

Real time alerts can be especially addictive and you have to remember there’s a point of diminishing returns. How much more insight are you going to get from 10 emails per day versus one in the morning and one in the evening?

Maybe this is a personal reminder, but I tend to get caught up in the numbers. For example, I’m fascinated by Google Analytics Real Time. I open it up and watch customers move around our site for far too long. It’s my version of watching TV. :)

But… you need to limit the amount of time you spend on measuring and analyzing. It’s important, but just relax and do what you feel helps the most during the week. Then measure it once at the end of the week and make a plan for what to test and adjust the next week. You’ll save time and sleep better at night.

About the Author: Josh Ledgard is the co-founder of KickoffLabs & SiftSocial.  He’s fanatical about his family, success metrics, and social technology.  You can read more of his posts on the KickoffLabs blog and his personal blog. Follow him on Twitter @evolvingwe.

  1. Thanks Josh for a great read! In my opinion, most people mistake at 2nd principle and they automate every report. This is what I also used to do some time ago. And other people (me included) don’t even know all of these principles.

  2. THANK YOU!!!!

    There’s a real lack of content out there that focuses on measurement and setting key performance indicators.

    Too many sites focus on every other aspect, but very rarely does anyone break down the importance of measurement to make it easy to understand for non data geeks.

    Now in your experience, what are your most important metrics, and how do you iterate to set new goals?

    • No problem Tommy!

      1. Our customers own conversion rates (since they setup landing pages with our platform)
      2. Our conversion rates and funnel.
      3. Our cancelation rates so we know the total lifetime value of a new customer. Which helps determine marketing spend.

      Josh

  3. Great article.

    A good way of not only establishing and monitoring KPI’s – but breaking them down to get deeper insights – is to segment by people who convert and also those who do not.

    Looking at average pages per visit, bounce rates, top content reports by segmenting converters vs. non-converters can be very insightful and gives you a direct comparison of how your website performs when people do what you want and when people don’t.

    You’ll be surprised at the difference in key metrics for each segment and will be able to pin-point KPI’s you need to increase or decrease to move the performance of non-converters into line with converters.

    Cheers,
    Ken

    • Agreed. I could have gone into that but we absolutely also split our funnel up by converts/ no converts to see the difference in audience.

  4. Thanks for this Article ! Good Work !

Comments are closed.

← Previous ArticleNext Article →