I was back with a startup I’m advising this past weekend discussing the importance of metrics. Or rather, more specifically, what you get with event tracking apps like MixPanel vs. Google Analytics.
So let’s start with an example from my office – tours at Atlanta Tech Village.

You’ve taken the tour before, but you’re here with 2 friends who have never been. Meanwhile, there are seven other tour guests to which two say they’ve taken the tour again, but they loved it so much they’re doing it again. 

The tour starts downstairs in the lobby before entering the rec room with game consoles, a ping pong table, shuffle board, and a kitchen. 

Then, the tour goes through the “hot desk” area before re-entering the lobby and going up a flight of stairs to the second floor with dedicated desks and some smaller offices. 

The group then goes up to one of 3rd, 4th, or 5th floors to see the larger office spaces. The tour guide is showing the group the kitchen as well as maybe an introduction to a startup who happens to have a door open. 

Then, the group goes up the roof to check out the sweet rooftop patio. 

The group may also then head all the way to the basement where the gym is. Then, the tour ends going back to the first floor – lobby.

Got the flow? Awesome. Let’s talk about this from Google Analytics and event-tracking with MixPanel (for example).
If tour visitors were tracked in a Google Analytics-like way:
  • You + the two other visitors who said they’ve taken the tour before would be Returning Visitors while the other seven are New Visitors.
  • How everyone came to take the tour – maybe referral from a friend like me, by an ad people saw at Farm Burger, or maybe they just walked by Piedmont Road.
  • What floors and rooms the group visited.
  • Great at telling you that most everyone came in through the lobby floor – the front door or the parking garage door. It can tell you some high level flows of how you traversed the building.
  • Perhaps a couple tour members got side-tracked and skipped a floor and met back up with the group. Or maybe a couple of them left mid-way through – exits.

Event-tracking with MixPanel would show you…

  • Three of us tour members are returning.
  • I’ve been to ATV dozens of times, and I took the tour three times.
  • I am Daryl and give you some contact information because I filled out some user information the first time I entered building.
  • Each room we enter (like Google Analytics).
  • Four of us on the tour played ping pong for 3 minutes before going to the next room. This is more detailed than Google Analytics may say we were in the ROOM for 5 minutes by telling you what four of us did in the room.
  • When we got to the second floor, two of us stopped to talk to a startup, and we also grabbed a couple drinks from the kitchen. Perhaps we also sat in a room so we could test out what it’d feel like to be a startup at ATV.
  • Two people who left the tour mid-way stopped by the bathroom for two seconds, and left the tour. (Maybe the bathroom was horrendously dirty.)
  • It was my friend, John, who hit the elevator button before we all went down to the gym.

… are you starting to see what event-tracking is? In this case, MixPanel is telling me more details of what happened on the tour – events. MixPanel is able to tie in user data because a few of us signed in from the beginning.

Google Analytics is able to quickly and automatically track much about our tour group’s visit. However, it’s still pretty high-level, and though it can be exhaustively tweaked to track a lot of events, it would take a lot of work to get the data and make sense of it all. Event-tracking apps like MixPanel are made for this stuff.
Google Analytics is a powerful tool that does a lot out of the box, and should be one tool for marketing insights. An app like MixPanel allows far greater insights for customer engagement, product roadmap, and yes, marketing.
Ah and another difference between Google Analytics and MixPanel… Google Analytics relies on cookies to associate site visitors as new or returning. However, if I stayed silent at the beginning of the tour and didn’t say anything, that’s like me hiding or starting a new “cookie” session. Google Analytics would count me as a new visitor. MixPanel could recognize me as a returning visitor in the system because I would sign back in as part of the tour.
Google Analytics. Easy set-up. Automatic high-level aggregate tracking. Insights via cookies.
MixPanel. Much more work to set up. Tracks the details (events) of visits and engagement at the individual level. Easier to build and determine funnels (drop-offs) of users as they move.
Hope this was helpful. Get trackin’! 

Having done sales and product management for a little while with startups including my own, I’ve realized how much “likes” mean little in the way of actual usage and conversion. 
I’ve heard plenty of times how prospects “like” a product or service, but then these leads go cold. Or a trial sign-up that never converted. What sounded like sure-wins were not at all. (This is why I get excited about a sale only after two weeks have passed since the check’s cleared.)

A lot can influence these gone-cold leads from poor marketing that gave the wrong expectations, lack of sales diligence and perseverance, competitive products and services, etc. 

You can sense a prospect’s buying interest by his emotion and his engagement in a conversation. I compare the “you have something here” or “I like how this does that” to the social media “Like” button. The term and the “feeling” is fleeting and seldom means anything of real consequence. 

Instead, as a salesperson, as an entrepreneur, as a product engineer, etc., we should be chasing love. We should be doing what we can to generate love for our product or service. Taking a few words from Sam Altman’s (President of renowned startup incubator Y Combinator) Startup Playbook:

Your goal as a startup is to make something users love. If you do that, then you have to figure out how to get a lot more users. But this first part is critical—think about the really successful companies of today. They all started with a product that their early users loved so much they told other people about it. If you fail to do this, you will fail. If you deceive yourself and think your users love your product when they don’t, you will still fail.

Love comes from actual engagement and use of a product. Love comes from solving a real pain point… a “hair-on-fire” problem. It drives prospects to buy and users to engage. It creates word-of-mouth marketing and social proof. Love is the foundation of successful companies.

Though, can you turn all those “likes” you get into love? 
On Tuesday, I shared a post on one of the greatest overlooked challenges by startups, and in particular, first-time entrepreneurs — considering change management
It’s imperative to consider the change management piece of any service or product — change management being the leading the transformation of the people within the organization. This is especially important with customers with multiple layers. 
Executives may be motivated to use a product or service for the benefit of the enterprise, but if the solution fails to integrate easily into the workflow of the “employees”, there’s a high likelihood of low adoption. With low adoption comes missed value which yields customer churn.
Change management involves a lot. I’ll take some examples from a SaaS perspective since I’m more accustomed to this arena — some key areas where a startup will surely fail:
  • Cumbersome UI/ UX. Said before many times and saying it again: UI and UX are now MUST-HAVES. Interfaces must be intuitive with clear calls-to-action. Bonus points for engaging UX will drive emotional evangelists.
  • Dramatic shifts from today to tomorrow. With Body Boss, we gave a lot of data and power to the coaches. We were able to chart out where players were after any session. However, most coaches weren’t accustomed to receiving that type of data continually. We didn’t guide coaches on how to interpret and make the data actionable. In many ways, we challenged coaches to absorb and use data that had never received before. Better practice would have been better reporting to drive focus and action to only the important data points.
  • Failing to consider the different consumers of a product. In consulting, you have the buyers who may be the execs representing some functional side of the business. Then, you may have a VP of Information Technology who had evaluate the software for security, for implementation, etc.. Then, there’s the end users of the system who may be front-line workers or managers who actually use the product. They all represented groups that consumes the product, and thus, it’s important to understand the benefits and risks of each group.
  • Not providing/ communicating (or very minimally) the vision and value of through the organization. That is, failure to share the value and get buy-in from the levels of the organization involved will, again, spur low adoption.
Obviously, there are many more ways to encourage failure through poor change management processes. Successful startups build a product and service that considers the people of the organization —their motivations, fears, and workflows.
One of the biggest challenges often overlooked by first-time entrepreneurs is change management. This was always a challenge when I was a supply chain consultant, and it was incredibly true as a founder of Body Boss.  

In consulting, I made recommendations, set strategies, and led implementations for transformations set by company execs. The transformations almost always had great intentions and great financial benefits that would otherwise seem like no-brainers to work. 
However, the greatest challenges weren’t the systems and processes to implement. Instead, the greatest challenges were leading and managing the change within the people of the organization. Different backgrounds, different skill levels, different motivations… everything impacted the people. The efficacy of leading change with the people dictated the success of transformations. Bain & Company cites “more than 70 percent of major change efforts typically fail” (see Results Delivery), and it’s largely due to change management.
Technology startups looking sell to businesses or consumers face a similar task in considering their products and services, especially in multi-layered organizations. In today’s world of quick-to-implement SaaS solutions, the effectiveness of change management can be a major component of user engagement and retention metrics (i.e. Day 1 returning users (“D1”), D7, D14, etc.).
When thinking about change management, entrepreneurs would do well to:
  • Consider the usage and implications to every layer of an organization that will use the product/ service — from executive buyers to core users
  • Communicate clear benefits, and ensure delivery of said benefits
  • Provide timely service in the event of missteps, bugs, or failure
  • Nurture adoption of the product/ service with all levels of users (getting started wizards, email notifications, other)
  • Create a simple, streamlined user experience (UX)
  • Understand and mitigate the risks to each [level of] user — considering your solution is from a new, questionably “viable” entity
I want to continue my post from Tuesday about the importance and value of instrumentation. Today, I want to share SaaS metrics that can be answered with proper instrumentation (operational and business).
  1. Cost of acquisition. This is cost to capture an unaffiliated buyer. Need to know the costs associated with closing this opportunity including marketing costs, engineering support, etc. For this metric, it’s important to track the flow and behaviors of a customer through websites, sales touches, etc.
  2. D1, D7, D14, D30 retention metrics. Here, D stands for “day” and the number refers to the number of days since a user first enters the system. This metric tracks the percentage of returning users to the service in D days –gauge “stickiness”.
  3. Open and Click-Through Rates of Emails. Many products these days have email engagement and nurture campaigns. Here, companies measure if users are opening these emails and, if applicable, are they clicking into a destination the company is looking for.
  4. Drop-off During Sign-up Process. Many products have multi-stage sign-ups which can deter and annoy users from completing sign-up. By measuring here, the company can quickly ascertain if the sign-up process needs to be simplified or be very valuable to motivate complete sign-up. If they never enter, they’ll never see the great product! (This, by the way, is why so many apps use Facebook, Twitter, Google login… plus, companies get personal data shared from those platforms.)
  5. In-App Engagement.This is a big bucket including what pages, tabs, profiles, features are viewed and used. You want to understand how users interact with the product – are they finding pages useful? Are features cumbersome?
  6. Customer Lifetime Value. Same concept of the revenue of a customer (or net profit) but extrapolated against the number of times a customer buys (subscription, multiple products, etc.).
  7. Churn.That is, what percentage of customers stop buying annually? Good annual churn for SaaS businesses according to Sixteen Ventures is 5-7%. High churn may point to poor value, mismanaged expectations, or an inherent problem in the product.
  8. Average Revenue per Customer. To be explicit, it’s total revenue divided by customer. In this case, revenue would naturally be weighted by where most of revenue comes from.

I get excited when a company properly instruments their products and services. It demonstrates tremendous maturity and understanding to recognize engagement data will drive confirmation (or rejection) of hypotheses, and thus, enables smarter business decisions.

What are some other metrics you find useful? How would you measure success in your company?
One of the chapters of my book is all about measuring engagement and implementing analytics – chapter aptly dubbed, “What gets measured gets improved.
The lesson is to “instrument” an application to gather data points of how users interact with the product. Quickly, you can assess if users are traversing all the steps of a Getting Started Wizard, getting stuck while building out a team, or exiting a page with above-average frequency.
We implemented Google Analytics at Body Boss which delivered anonymized data at an aggregate level – useful, but doesn’t give finer perspective into engagement. More powerful instrumentation would including capturing “events” to every actionable UI element (i.e. button), page, etc. For example, in Twitter, events would include when a user navigates to a user’s profile, scrolls down, Likes a tweet, and then searches for another user.
There are many fascinating tools for instrumentation on web pages, in-app uses, and beyond. Here are a few I’ve recently got to play with:
  • Fullstory– records behind the scenes how users interact with a site or app that can be replayed later. You can see where a user hovers his mouse, scrolls, stays on some block of text, etc.
  • Unbounce – taking A/B testing to multi-variate testing for landing pages. Quickly set up a landing page with multiple variants, and Unbounce automatically directs visitors and tracks conversions.
  • Pardot – Easily send automated mass messages personalized to recipients based on where users are in the sales funnel. Tracks users from first site visit and beyond for nurture campaigns
  • MixPanel/ Intercom – Very different in how each operate, but the feature I liked most was being able to trigger specific messages (more granular than Pardot) based on user interactions. High-degree of control by building out event-driven rules and trigger notifications.
  • Kevy – Marketing automation for ecommerce stores. Slick tool to understand consumer behavior and enables stores to better market to consumers by offering coupons, messages, and the like based on rules and triggers.

In gist, there are lots of tools available for instrumentation with overlapping features. It’s fun learning about these tools now, and dreaming of how great these would have been at Body Boss. Though, several tools didn’t exist three years ago… inherent problems don’t change, but solutions do.

What are your thoughts on instrumentation? What tools do you use?
(Source: http://s306.photobucket.com/user/paawkx/media/mario-foresight.png.html)
Rebekah Campbellknew at the beginning that the original plan for her tech startup wasn’t going to work. Yes, even before launching. Campbell is the Founder of Posse, the app where the community of your friends share recommendations and favorite places.
Campbell recently shared her experience on the New York Time’s blog – “When Plan A Doesn’t Work”. She had received funding for her loyalty app, to put it simply, but from the get-go, she knew that it would suffer. In fact, it crystallized as she forgot to download the loyalty app her hair stylist told her about after her appointment. She had focused so much attention on the merchant side, but didn’t think much around user engagement.
In many businesses, products and services actually cater to more than one user beyond your actual customer. Campbell’s experience is reminiscent of Body Boss, and how the two user groups we had to cater to – coaches/ trainers and players.
Some key take-aways from Campbell’s piece:
  • Step Into the Shoes of Your Users. Campbell heard about a loyalty app for her hairstylist’s salon at her appointment, and said she’d try it out. However, she forgot to download it. How would users of Posse download and start engaging right off the bat if she, like everyone else, forgot to download another app so quickly? (See “Eating your own dogfood and imitating the late Steve Irwin – Why would you do that?”)
  • Early Customer-Partners Are CRITICAL. Campbell was able to convince so many people prior to launch 2.0 to come in for focus groups, 1-on-1’s, and other tests/ interviews over and over again. That’s crazy amazing. To have a list of early customers/ users bought in to help you iterate and figure out what works is so incredibly valuable so you don’t build something no one wants.
  • Appeal to Those Who Like You. After focus groups, Campbell and team not only saw distinct behavioral groups, but that not all would be as receptive to Posse as others. Thus, she honed in on those groups that would be more inclined to engage with Posse. Save time, save money with focus.
  • Personalized Recommendations is the Future. The likes of Amazon reviews, Yelp reviews, etc. are great to kind of discover new places. However, how much credibility do you actually give some of those reviews? Instead, you more often than not ask those around you for recommendations of a good stylist, the places to go on vacation when in Rome, or where to go for BBQ in Atlanta tonight.
  • Be Beautiful. Be Simple. Be Novel. Everyone and their mothers are getting into tech these days. I regularly do a purge of apps on my phone if I hadn’t used certain apps after 2 weeks. With all that’s out there, it’s getting more difficult to stand apart. I found the following to be one of the most powerful nuggets in her post: “Like a lot of people, I’m lazy. If I’m going to try something new, it has to be so useful, so fun and so original that it blows my mind. Otherwise, forget it.

Of course I would love and appreciate this post – it was published on my birthday! Well, okay, it was obviously for its content. At Body Boss, we realized early on that we had two major user groups with very different inclinations for technology in coaches and players. We didn’t quite address this as quickly as we would have liked, and that could be a factor of our bootstrappy-ness; whereas, Campbell could iterate quickly with full-time employees with venture capital.
Campbell’s Posse 2.0 launched in March 2013 with over 70,000 users and 40,000 stores globally (at the time of publish). Looks like they’ve got some good momentum, and I’m hoping they go far.

What are your thoughts about Campbell’s article? How else could Campbell have iterated quickly to find the right product/ market fit to address customer engagement?