02 10 / 2013
After nearly three years of leading acquisition for a SaaS startup, I’ve learned that all acquisition efforts are best when measured against three expectations: volume, quality and efficiency.
Volume - This should come as no surprise. Startups need to acquire customers in volume, month after month.
Quality - Everything from the top of the funnel to paying customers is measured against quality. It doesn’t matter if you get a million visitors or 50k customers if they are shitty and don’t stick around. Common SaaS metrics like conversion rates, LTV, revenue churn and ARPU will be heavily impacted by low quality traffic & customers.
Efficiency - Acquiring customers cost-effectively according to your company’s unit economics is essential. The most critical aspect to creating efficiency is having solid analytics in place to understand which channels, campaigns and keywords drive actual revenue.
If you are constantly evaluating against volume, quality AND efficiency, it becomes much easier to create effective growth. You avoid paying for loads of low-quality traffic/leads, driving your CAC through the roof, bringing growth to a halt and many other bad scenarios.
30 4 / 2013
This article was originally posted on the Leanometry Blog.
Whether a product starts out in beta or go straight to a public launch, startups often spend a good chunk of time trying to gain feedback from users on the new product. This feedback is helpful and sometimes critical in understanding what is missing or what shouldn’t be there to make it really work well for the user.
However, what the emails, phone calls, chats and surveys won’t tell you is how the users are actually interacting with your product. Where are they clicking, how much time are they spending on each page/section and what is the typical ﬂow of a user once they’ve gone beyond exploring the site/application? In order to have context and hard data to compliment the qualitative data received, you need to implement speciﬁc tracking and reporting within your product.
With the free and seemingly-ubiquitous Google Analytics, you can get all of this quantitative data and more. Below are three suggestions to help you get started:
1. Event Tracking
Make the effort to add a line of HTML to the links, buttons and videos in your product and the data you get should be pretty revealing. When you can see where the most clicks are occurring, you can guide changes in the application to give certain items more prominent real estate than others.
For example, if a new user is presented with a features tour modal, you may ﬁnd that over 50% of them are clicking ‘Skip This’ instead of ‘Next Feature’. This data can be used to test which feature is shown ﬁrst, the design of the modal and even the timing.
<a href="#" onClick="_gaq.push(['_trackEvent', 'Videos','Play','Homepage Testimonial']);">Play</a>
Within your product, ﬁgure out what actions you think are most important for users to be successful and then create goals around them. It could be as simple as meeting a certain threshold for visit duration or as complex as a speciﬁc group of click events. By having these goals in place, you will be able to see how changes within the product affect your top goals.
3. Content Reports
The ﬁrst two suggestions have been more granular, so to back out and see a bigger picture, the ‘All Pages’ content report is great. As long as you have GA tracking code on all your product pages, you should be able to open the default report and start to understand what sections/pages users are spending the most time in.
As you look through the volume of pageviews per page along with the standard metrics provided (Time on Page, Bounce Rate, Entrances & Exits), you can start to get a rough picture of how users are interacting with your product at a macro level.
Once you ﬁgure out your tracking goals and get everything set up, you can start to weave together your qualitative and quantitive data to get a better picture of how users are interacting with your product, what they want and how your changes may potentially impact your goals. As you make changes and updates to your product, the data will serve as a good baseline to understand user behavior and again, give context to the feedback you receive.
If you have a new product or are nearing a launch, I hope you consider setting goals and tracking product usage to see the full picture and make intelligent changes to your new and growing product.
12 4 / 2013
I hear a lot of varying opinions on the bounce rate metric in analytics. Some think the overall bounce rate is a key metric in site performance and others think it is completely worthless. I’ve gone back and forth over the years and ended up landing in the middle, believing that both sides are wrong and right.
Why Overall Bounce Rate Is Dangerous
The overall bounce rate is good to get a quick read on how people are engaging with your site, but it is an average of all pages and that is precisely why you can’t rely on it as a KPI.
Simply watching the overall bounce rate will only show you incremental change. The more pages on your site, the less your bounce rate will change if a small group of pages spike from 40% to 90%. You aren’t going to jump in to find problems if you see your bounce rate go from 43.12% to 43.19%, but this may result in several revenue-driving pages performing poorly.
Even if you do want to dramatically improve your overall bounce rate, the way you will do that is by drilling down into your landing pages to find areas for improvement. So, get into the details and don’t live or die by the average.
Why Ignoring Bounce Rate Is Dangerous
While the overall bounce rate doesn’t give context, it is crucial to be tracking bounce rate for specific landing pages & campaigns. If you are spending money on traffic, you should understand how much is bouncing. Changes in keywords, targeting and ad content can change the experience for that visitor and the bounce rate for each landing page is an excellent signal for understanding if the change was positive or negative.
Outside of paid traffic, there are still going to be a lot of pages on your site that will be driving leads/revenue organically. If your bounce rate goes up on these pages, it is likely that your conversion rate will go down and leave you with less money at the end of your funnel.
I would recommend identifying your top revenue generating pages and creating a custom dashboard that shows stats for those pages so that you can use bounce rate as a signal of possible performance issues. The pages driving the most value for your site should be viewed frequently and easily accessible.
How to Approach Single Page Sites
If you have a single page website or an orphan landing page, it may appear that your bounce rate is incredibly high. Unless your page is strictly informational, there will likely be one or more actions you want the visitors to take.
If you add Event Tracking to any of the buttons, links or videos on that page, Google Analytics will count those actions as an interaction. This will prevent anyone with an interaction from appearing as a bounce and help you understand how each channel is truly performing.
The Only Way to Use Bounce Rate
Use the bounce rate metric for context at the granular level. Don’t use averages and don’t make it the only metric you optimize against. Let it signal what visitors are expecting to see/do on the page and align your messaging to meet that expectation.
15 1 / 2013
Stretch goals are a very common thing. I set them for others and have them set for me. Over time, I have discovered that stretch goals alone are not effective.
The Motivation Problem
The whole premise of setting stretch goals is to aim so high that it inspires you or your team to do better than if lower goals were set. However, when you do this regularly, it can create an expectation of certain failure and be more discouraging than motivating.
People like to succeed and achieve goals and purely working against stretch goals can keep them from ever getting the satisfaction of accomplishing those goals.
The Business Problem
In using stretch goals for business, the idea is that you set goals so that if achieved, you would be doing insanely well. If you come close to them, you’ll be doing great. The problem again with only using stretch goals is that regardless of that number, you still have an expectation of what is reasonable.
"I would love to get x, but I will be happy with y."
When expectations are not coupled with stretch goals, there is a risk that those working toward the goals will not know how to achieve the bottom line of what is acceptable.
By communicating both stretch goals and expectations, people can still aspire for greatness while ensuring that they still reach the numbers/levels necessary to continue succeeding. For those working toward the goals, there will be a greater sense of accomplishment and more motivation to hit the stretch goals. Everybody wins.