Wednesday, December 4, 2013

Modern Digital Analytics Tools

Here's the original copy of an article I wrote for You can see the published version here if you like. It's based on my presentation at the Seattle Digital Analytics Association which I have to say was a whole lot of fun. 

A new generation of digital analysis tools are arriving. A generation where speed is king and customer data is united. If you have been buried with reports and analysis from the typical tools then take a break and check out these six modern analytics tools. 

Heap Analytics
“Capture everything” is the theme of Heap Analytics. Similar to tag management a small snippet of code is added to the site. Then a flood of data begins to roll into the system. This data comes in raw in the form of an event feed. Every click, tap, swipe and page is captured. Using the interface these actions are labeled “Facebook Like” or “Twitter Follow”. Super Events can then be created which are a combination of events. In our example “Social actions” would be a super event containing “Facebook Like” and “Twitter Follow”.

If data was not captured “right”, simply re-label it. New pages can be shipped anytime without waiting for an analytics QA. Once the events are defined, then the real magic of Heap takes over.

Unlike other analytic tools which group data fundamentally by page view, Heap organizes by visitor first. If there is no Visitor ID specified, it uses its own Visitor ID. Then when segments are defined such as “People who partially complete a form” the individuals who make up the segment can be listed. Further, the event stream related to each customer is attached to the ID so it’s possible to easily see the paths each person followed as they fell into your defined segment.  

Heap has a sliding pricing scale based on unique visitors. 500,000 unique visitors runs approximately $2000 a month. is part of a new generation of tools designed to connect all your customer data and then turn around and power other marketing tools with sharply refined segments. Lytics can start with a small pool of data such as email addresses or web logs, then using a proprietary matching technology, it adds color to the customer record using external data sources from Rapleaf, Facebook and keyword search data. The end effect is the creation of a “gold customer record” using an ever-sharpening picture of your customer.

Using a series of rules, very fine segments can be created that use triggers to send emails, post messages or otherwise reach out to the customer. The customer record includes active internet times, active locations, active devices, subjects of interest, demographic and psychographic data.  

Lytics pricing starts around $5000/mo. for the initial model.

Nectar Online Media specializes in “hyperpersonalization” driven by the unification of social media streams. Nectar also unites data to form a holistic view of the customer, however the goal is execution of speedy triggers designed to nudge customers into action at exactly the right time. For example, Ryan browses pages on your site showing the various models of cameras you sell. He spends a lot of time on one specific model. Then he leaves but doesn’t buy it. Using data from Ryan’s Facebook stream, you see he has an event coming up… he’s going on vacation to Belize. So you post a coupon for $100 off said camera in Ryan’s Facebook newsfeed. Then you use remarketing techniques to advertise the camera and accessories to him as he browses around the web. This is hyperpersonalization. 

Nectar pricing was not available.
Right or wrong, beautiful visual presentation of data can control how much it is read and consumed. But few graphic designers have analysis knowledge and few analysts have design skills. Enter is an online system with ready-made infographic templates as well as customizable graphs, word clouds and icons. You can also upload and embed your own graphics and video.

The system is very user-friendly. If you have data ready to go then in about 5-10 minutes you can produce your infographic. Most of that time is exploring the various customization options. is a static system so once the graphic is posted, you cannot drill and dig into the numbers. It’s also not designed to take in huge volumes of data. But for meaningful summaries designed to get attention, it’s exactly right. is free but if you desire more templates, there is a pro version for $12/month.

Insight Rocket
Automated multichannel storytelling is the way to think about Insight Rocket. This tool combines the reporting strength and flexibility of Tableau with social commentary. This is particularly useful in larger companies where there may be multiple teams, brand and commerce for example, that define the numbers in different ways. Ideally, everyone should be using the same metrics but this is not usually the case. The teams need a way to connect and agree internally before data is surfaced as “truth” to management.

Insight Rocket also has a strong data integration team to help break down data silos in the first place.  So once data is combined and the Tableau reports are running, the analysts are free to dig into the data. Once an insight is found, it’s published just like a story and can be fed into email or an intranet system. Questions can be asked and answered right inline with the story.

Insight Rocket pricing ranges from $2000 - $10,000 per month depending on data sources and required support.

Beyond Core
Beyond Core represents a new frontier in data analytics. It combines true data science methods with machine learning and automated video. There are a collection of features on this tool for beginning analysts as well as seriously advanced analysts. For beginning analysis or non-analysts the guidance feature is nothing short of amazing. After data is loaded into the system, the machine takes over charting and plotting the most interesting data *and then* providing an automated, intelligent, animated voice over that describes in 2 minutes the most salient points. Since approximately 95% of the population cannot easily interpret analysis, this is an extremely useful tool. Ask any analyst who has explained what the data means multiple times to the same stakeholders how nice it would be to send this out automatically.

Beyond Core also opens its platform for advanced analysis operations. Since the system is Hadoop based, it can process millions of data rows in hours, not weeks, without sampling.  As the data is loaded the analyst selects what column the system should optimize for. Then the number crunching begins to show what factors are really driving the numbers. If an obvious answer comes back (i.e. the number of uninsured patients is driving up hospital costs) simply customize the analysis by throwing out that group and run it again. The second time, the deeper correlations show up, an operation which usually takes weeks to re-run. Further, because hundreds of variables can be included, any initial variable selection bias is factored out. 
Beyond Core is a self-service tool and pricing begins at $500 per month.

The big data revolution is driving a lot of change in digital analysis speed and tools. Senior management does not know it yet which places analysts in a wonderful position of looking like a hero.  

Monday, October 28, 2013

The Coming Consumer Revolution

You might not realize it but there is a huge shift happening in the world of digital analytics. For years people in this industry agonized over visits and page views and extrapolated behavior from these markers of engagement. This made it difficult to understand consumer behavior because it was so fragmented. But all this is about to change. 

Companies have always collected consumer information. Demographic marketing areas, zip code data, and income levels – none of this is new. What is new is the sheer amount of online "big data" swirling around that is just now starting to become connected to individual consumers instead of households or zip codes. 

Before, offline personas did not match online behavior, but that has changed. Companies know much more about you than you realize. So, what does this mean? We’ll see this the effects take shape in various ways:

1. Consumers will Game the System.
If you've ever shopped on an ecommerce site, put an item in your cart and not bought it in hopes of receiving a coupon to help convince you to purchase, then you've gamed the system. My husband recently had a similar experience on Amazon when he put an item in his cart, went to a competitor to check pricing and, suddenly the Amazon price dropped to be lower than the competition. As corporations try to more tightly control consumer behavior, consumers will quickly figure out these controls and use them to their advantage. 

2. Consumers will become stronger Partners.
I am not a big fan of sharing my data. In fact, I'm probably more resistant than most because I know how valuable it is, so I typically won't give it away unless the exchange is worth it to me. However, when a site like BlueKai presents me with all the information that's been aggregated about me, my desire is not to blow it away, but to reasonably correct it. My theory is that if I correct it, good things may come to me. I want Nike to know I'm a fan of their shoes but that I think their women's clothing is unoriginal. I want K2 to know I love their skis but I haven't bought a new pair in several years. I want them to know this because I want to be treated as special by these brands. And why not?

3. Companies will be more selective.
In a unified data world, a company might realize that their best buyers with the most lifetime value are actually in select zip codes. Why run specials in all geographic locations when there is an opportunity to be more surgical? Or even better, why not just specifically invite these high-value consumers to come in? An early example of this can be found at casinos who mine their customer databases for the high spenders and then attach personal sales reps to each one to individually call and invite these folks to come back along with free rooms, dinners, etc. Could you imagine being such a valuable consumer of say, United Airlines travel that they would pick you up and personally drive you to the airport? As companies slice the data, I believe "surgical service" for high value consumers will become part of the marketing game. 

4. Consumers will be pressured to stick with brands.
Once companies have identified you as valuable, and started to sweeten the benefits you receive, it will become harder to switch, assuming the consumer continues to find value in these benefits. And that will put more pressure on companies to keep you until they reach a point of optimization where the value you bring in (say your purchases over one year) does not exceed the cost of keeping you. Then the pressure switches back to the consumer to stay in the high-value zone or accept fewer benefits.  

Companies have more to gain than they realize by becoming more transparent with their consumers about how they market to them and what they know. Consumers have more to gain as well. Beneath it all is the overarching need for transparency and that's only a matter of time. 

Friday, October 11, 2013

DAA Seattle Symposium 2013 is Coming!

I hope to see you all at the DAA Seattle Symposium 2013 on November 6thThis year’s agenda has been extended to a full day to accommodate more speakers than ever before with presentation from Microsoft, Apollo Group, Bing, Adobe, Symatec, CardinalPath (that would be moi!) and many other leaders in the digital analytics space.

Space sells out quickly so I recommend registering as soon as you can. The event website and agenda can be found here:

Registration is only $40 for DAA members and $95 for non-members.

I look forward to seeing you on November 6th and please spread the word!

Tuesday, September 17, 2013

Personal Leading Measures: How Non-Verbals Give You an Edge


In digital analytics we have a concept called ‘leading measures.’ These are the early cues used to predict behavior. So, for example, a person looking at a product detail page may be leaving an early clue that they plan to buy the product. The same concept applies in business except it is called non-verbal communication. 

Understanding the messages you give and receive non-verbally is critical for all leaders, but especially for executive women. Recent research by TED speaker Amy Cuddy shows, in a quantitative way, that a person’s physical posture and expressions actually shape how they feel, and correspondingly, who they are. If you care about being a leader (regardless of gender) you must watch this TED lecture immediately. It might change your life; it's that good.

One thing Amy touches on is executive presence, or put more simply, physically expanding into the immediate space without crossing the “threatening” line. If you've ever had your boss put his or her hands behind their head, elbows sticking out while talking to you or seen someone put their feet up on the desk, then you've seen this kind of expansion. I do make a point to be aware of my own posture in meetings (even though I rarely put my hands behind my head like that.) I do sit up tall or stand to speak when I can. I also make a point to watch the posture of others – and doing this is without a doubt, business strategy.

Reading non-verbal cues is the next best thing to mind reading. It's like a poker game where combinations of "tells" repeated in clusters give you a clue about intent. Just like leading measures, it's a predictor of behavior. Remember, viewing the product detail page does not mean I will buy the product in the same way that observing non-verbals does not necessarily mean something will happen. So if I see two co-workers feet pointing at each other during the office Christmas party, which is an indicator of romance, it does not mean they are dating. Read more about these tells from Vanessa Van Edwards's great site.

What should you do now armed with this knowledge? Getting a clue early allows you to proactively manage it kindly and appropriately as a powerful female executive. Amy has some follow up research that appeared in the July-Aug 2013 issue of Harvard Business Review about the need to connect first, then lead. Simply put, when we display warmth first, then competence, the result is influence.

Bottom line, as you strive to become the confident, intelligent executive women you know you are, make good use of your personal leading measures to 1) use non-verbal preparations to present yourself well and 2) give yourself a strategic edge by monitoring the non-verbal data from your surroundings.

Thursday, September 5, 2013

Do You Have Good Quality of Life at Work? Here's How to Measure It.

This is my original, less edited version. See the one that appeared in Fast Company here

Many folks have asked me recently about my career move from Semphonic (acquired by Ernst & Young on March 31, 2013) to digital intelligence agency Cardinal Path. On the surface, it's easy to point to the larger company environment and the acquisition and say, of course. Yet this was actually a very tough decision. No employer is perfect and I feel I know this well having been one. So what if we could measure quality of life against any employer? Here's how I would, and did, measure it before making the call.

Quality of life at work is heavily correlated to time. We are all such Bisy Backson's so busy, busy, busy. At the beginning of the year I made a resolution to spend less time at work. I really wanted to keep it to the 40-45 hour range. As Americans we wear the number of hours we work in a week like a badge of honor. It's easy to overhear "Oh, I worked 60 hours last week. I'm so tired. Yeah, I worked all day Sunday and part of Saturday too." The statistics on this subject point out that we're not really working at this point. We're just becoming bit-shoveling, mouse-moving zombies. See this story at Inc. Magazine on Why Working More than 40 Hours a Week is Useless.

So my hours are about quality and value. When I spend 40 hours, it's 40 quality hours moving businesses forward, delivering work, checking in with my team, and taking a few breaths to think "what else could we do? what's new in the field?" It's just enough to spark product innovation and creativity which is fun. So in measuring quality of life against an employer, I look at how much time does it take to do a job I feel good about? At the big firm it took me about 55 hours to do what I could previously do in a 40 hour week. For me, this was simply a trend in the wrong direction.

In the corporate quality of life measurement framework, measure time as a ratio between busy hours vs quality work hours. If that's confusing, think of it as things you like to do vs stuff you do not like. At 1.0 you'd be fully optimized. Mine was more like 65/35, meaning 65% of the time I was doing things I felt had little value.

Ability to Change
A big ship, with all its power and luxury, is still hard to turn. Back in my start-up days, the ability to move faster than the big guys was often our sole method of survival. Fewer contractual issues, fewer approval processes, eye-opening innovations, and blinding speed. Exciting and addictive, the ability to move at speed with an industry is critical to digital intelligence.

Part of moving at speed is having the technology, the apps, the access to tools that enable change. Technology is a good proxy for a company's ability to change because the ability to use it causes constant innovation and flexible thinking (Windows 8 anyone?). At EY the technology was approximately 3-6 years (yes, years) behind. And while change was occurring, it was not leapfrogging ahead but simply moving to more like 2 years behind. The sheer force it took to make change occur was incredible. For me, the foreshadowing of effort to simply catch up while my industry charged ahead meant I would eventually lose what I loved most: innovate and go.

In the corporate quality of life measurement framework, measure ability to change as a ratio between existing in-house technology vs current industry technology. In this case I used 1 for current year tech and incremented +1 for every year behind. I then tacked it to a general 1-10 scale. My situation came out at 6/1, meaning the ability to change as measured by technical innovation was a 6 on a scale of 1 to 10.

Companies are like people. They all have specific goals and motivations. Sometimes we are aligned, for example, we both like public speaking. And sometimes we are not, for example, you like cutting costs and I want you to pay for my lunch. The trick is to sort out the minor stuff from the major stuff. And remember, it's rare to have a 100% fit with a company, a friend or a spouse. If you do, wow. Pat yourself on the back.

To measure alignment, list all your gripes of working for the company and then plotting the density of the items on your list as a scatterplot as follows. The x-axis would be how much you care about it (left = not much, right = a lot). The y-axis would be how frequently the issue comes up (low = not much, high = a lot).

Then divide your chart into 4 equal parts and label:

  • Meh - Items in the lower left corner probably didn't even make your list. These are things that do not happen often and you don't care much about them. 
  • Annoying - Items in the upper left corner probably did appear. These are annoyances which probably have a good frequency but you don't care about them as much as other items. 
  • Accommodations - Items in the lower right are accommodations you may be willing to make. These are things you care about, but they do not happen that often so they could be things you could live with.
  • Gotchas - Items in the top right are the real deal breakers. These are items you care a lot about and they happen pretty frequently. 

Note the density of the dots. If you have a lot of left leaning items, maybe you could make some small changes to get over it? If you have a lot of right leaning items, you might want to think seriously about whether your personal goals are heading in the same direction as the company.

You can also do this with positive items. In that case, I'd group them into the following four areas: Meh (stays the same), Annoying becomes perks, accommodations becomes benefits, gotchas becomes jewels / things you treasure.

Deciding whether to leave your company or stay should never be a flip decision. It can be difficult to sort out the larger issues from the minor annoyances but using metrics allowed me to clarify my view and make the right call.