6 keys to a killer career in web analytics

Every day, I become less of a dumbass, usually as a result of reflecting on what happens when I am one. Here are a few lessons I have learned that may be helpful to you, whether you are just getting started or if you’ve been doing this for a while. I hope you’ll add your own lessons to the comments.

So, here are 6 lessons I feel, if I had learned earlier, would have really helped me out.

1. Learn the tool(s).

Duh.

Get training. For Omniture, you can’t do better than Keystone Solutions. For Google Analytics, you can talk to us, or any GACP. Learn the way the tool is laid out, how to use its advanced features, how to get lost going down a rabbit hole but still know what you are doing.

And then this (warning, you might not like this): take 2 vacation days and spend it getting completely lost in the tool. Click on surprising data to see why it’s doing what it’s doing. Click on unsurprising data to gain an appreciation for the surprises that probably lurk underneath. The vacation days will allow you to get away from everyone else, get away from specific requests, and focus on getting lost in the data and the tool’s capabilities. You aren’t going to break it. Learn all of the nooks and crannies. Click on everything you can possibly click on (maybe not in the admin side of the tool, watch your clicks in there). Make a list of questions, and get them answered by someone who really knows their stuff.

Finally, learn more than just the basics of implementation. This will do two things. First, it’ll help you understand how the data is gathered and why you might see odd things happening. Second, it’ll help you understand the pain involved in implementing and how “something simple” really isn’t. Use this understanding to build a relationship with your IT team, or even better, justify a tag management system (Satellite, Ensighten, Tagman, Brighttag) that will get you focused on using the data, not getting it. Focusing your time on getting data is not a good thing.*

* Don’t give me the argument that without getting data, there is no data to analyze. Without celery, there is no ants on a log. Without rain, there is no celery. Without evaporation, there is no rain. Without glaciers and the ocean, there is nothing to evaporate. Without hydrogen, there are no glaciers. Without the big bang, there is no hydrogen. Give me a freaking break. Which brings me to #2:

2. Focus on the output

Particularly, when presenting. Know your audience and deliver first, before anything else, a message that gets their attention. Your process is your content. Your conclusion is your headline.

Maybe conclusion is a crappy word, since it’s usually something you figure out a the end. But rather than making your presentation a playback of the process you went through, reverse the order and start with what you found out at the end. Usually, this is just the “hook” that you need to get everyone’s attention, get them asking questions, and help them arrive at the same conclusion you did.

I’ve found that when I start by walking through my process, the audience knows they are in for a long, boring meeting where I don’t get to the point for quite a while.

3. Focus on YOU

YOU can answer questions. YOU can get information and numbers. YOU can tell executives what people think about the web site. Not tools.

Yes, the tools provide this information, but YOU put it into terms (a la “focus on the output”) that the business can consume. SiteCatalyst was not custom-built for Best Buy. OpinionLab is not only installed on Williams-Sonoma. These tools have to be built to handle any web site and every business model. It is you who is valuable; you are the one who adds context and a sense of reality and applicability to the data.

Without tools, you can’t answer questions. But with tools, it doesn’t mean you can answer questions effectively.

How many people saw Tiger Woods crush the competition at the Masters and then went out and bought the clubs he used to win? Did those clubs work for these people? People think they can buy a better game. They think the tools will get them on the podium. Unfortunately, my golf game proves without fail that this is not the case.

Try to get people to ask you the question, not ask you to get data. If they do ask you for data, show them what they asked for behind a page that you put together. Try to illustrate on that page why the request for the data could have led them to the wrong conclusion (if that’s true) or didn’t tell the whole story. You don’t have to do it in an overt way, just show that the human is where the value comes from, not the tool.

Once you realize that you are the bomb, it’s time to…

4. Focus on THEM

When you are delivering your powerful message and wrapping that in your own brilliance, remember that even if your conclusion is the same, you will see that people from different departments will grapple with your ideas in radically different ways. The same sentence read by or spoken to a CEO, a CFO, a CMO, a Paid Search Manager, and a Usability Engineer may mean different things. Some may feel the message helps them. Others may feel threatened by it. Some may see savings and cost reduction while others see upside. Some will see increased productivity while others will see their fiefdom shrinking.

I have made huge mistakes in my work by throwing an “unframed” idea out to a mixed group of stakeholders or by framing it the same way to all stakeholders. Huge mistakes.

By learning what goals and motivations each of the players has, you can show each player how an idea benefits them. If it doesn’t benefit them directly, you can help them understand how it benefits others in such a way that it may warrant sacrifice (which may be repaid in the future). You will also understand their goals and motivations for the future, allowing you to be on the lookout for ideas that do benefit them, which will build a tremendously successful relationship.

If you see salespeople or other charismatic people who are seemingly able to do this on the fly, don’t despair. There is no gene for this: these people spend a lot more time than they let on trying to see things from other peoples’ perspectives. Charismatic people work hard to understand others’ wants and needs so they come into these meetings prepared to speak in a way they know will be successful. By giving others’ perspectives some forethought, you literally stack the deck in your favor, but it does take conscious effort.

5. Be the closest thing to a graphic designer possible

If a picture is worth a thousand words, why do so many people just say, “wow…,” when you put something truly amazing in front of them?

Some of my most successful presentations had just a single chart on the first slide. I make an introduction, and say, “Here is what we need to figure out today…,” change to this slide, and just watch the faces in the room.

A book written by the company that helped Al Gore design his famous “Inconvenient Truth” presentation shows a chart of two different kinds of companies. One line (red) shows the market cap of companies who have design deeply-ingrained into their company and culture. The other, blue line shows the rest of the companies in that exchange (a UK exchange). Over a period of about 7 years, the companies who had design at their core had grown by a factor of 2x compared to the companies that did not.

wow…

When you are able to convey ideas clearly, your customers know what your product is, what it does, how it works. When you can convey your thoughts clearly, and this usually happens graphically, people “get it” instantly. They want to know more. And they trust that when they ask questions, those questions, too, will be answered in a clear way.

Looking at companies and people who do not communicate through effective design, on the other hand, you may feel that interactions with these people are cumbersome and unproductive, that your questions are answered in round-about or indirect ways, and you are generally less-likely to interact with someone or see their value clearly.

Personally, I’m about as good of a graphic designer as Hellen Keller. But I try as hard as possible. I try to strip things down to their most basic elements, show them to people, get feedback. I try to be a student of effective communication. I can assure you it is a skill that can be improved.

6. Focus on reaction, not perfection

If you could answer a question 10 times (instead of just once) on Jeopardy, what would be your strategy? If it was me, I would answer as quickly as humanly possible with a semi-educated guess. What if Alex told you if you were getting “warmer” or “colder” with each answer? Yeah, I’d be jumping all over that buzzer to ring in.

Online, this is exactly how it works. Yet we are obsessed with trying to answer the question “right” the first time. We believe that analytics is a practice designed to help us make better decisions. I disagree.

I think that digital analytics is best used to help us make more decisions, more quickly. I would just about wager my life that if I can make 5 decisions in the time it takes you to make 1, I will get a better net result almost every time. This is the premise of A/B and multivariate testing, but even with tests, people criticize the “loss” in the test variants that fail to find upside or in a test where all variants produce inferior results.

“Getting colder” is the same as “getting warmer.” Remember that.

When we use analytics to make better decisions, it’s like when a shop owner tries to learn everything he possibly can about how to organize his shop, stock his shelves, hire help, and get it all just right for opening day. Imagine the OCD shopkeeper who delays his store opening for a month because he keeps getting little bits of information, reading just a few more articles in hopes he’ll get it just right on opening day.

Then the store opens.

Which period of time do you think taught him more about his customers and how to run his store? The months before opening, or the hours after?

When we use analytics to “open the shop early” and get ideas out into the real world, we can learn quickly and react. In the time it would have taken us to learn what should work, we can learn 5 times what does work.

There are exceptions, but to me, this is the rule.

See the light

What is the eventuality of your career? Where is this taking you? In 10 years, when someone else is putting the reports together, delivering the types of presentations you are delivering today, where will you be?

I don’t know the answer to this question. Not by a long shot. But my gut tells me that what we are doing today will train us for some pretty impressive roles in the future. And if that is the case, my gut tells me that these roles will get farther away from the kinds of metrics we work with today and closer to the kinds of metrics and questions that Wall Street works with every day: How efficient is your company? How is it growing? What is the promise intrinsic in the business’s plans for the future? What evidence is there that new business opportunities will pan out?

Those are things we are probably pretty good at today, and will be incredibly good at farther down the road.

Your turn

This has just been my experience and a few things I feel help me. But what about you? Share your thoughts in the comments.


No Trackbacks

8 Comments

  1. Good post, Evan. There is no question that this is great advice. I am adapting to life as an analyst right now, and I have already come across some of these points; others are items I hadn’t considered but which I can clearly see will benefit me. My request as a reader is that you provide some additional depth. For example, “being the closest thing to a graphic designer possible” has obvious benefits in the work that we do with other teams and management. But how exactly did you do it (by the way, I have seen you present; your design skills are better than you believe)? How should we do it? I’m not asking for a list of resources, either; with each of these points, I’d like to hear about your experiences. Hammer home this great advice with anecdotes and extensible examples of both success and failure. Otherwise, career advice often seems too abstract and/or too difficult to be acted upon. So, I guess I’m asking you to spin this off into a series of posts that deals with each of these points more deeply and more personally. Give us homework! Challenge us in very specific terms!

    Anyway, thank you for continuing to share your insights with us. I know I have benefited tremendously from them.

    Posted January 19, 2012 at 3:32 pm | Permalink
  2. Excellent work Evan! I too believe the true analysts out there (as opposed to those doing reporting) will end up in radically different places than they can imagine today. It’s about how the brain works and natural curiosity, not the tools.

    Web analysts are honing the skills they need for “Business Optimization” right now.

    Most of these folks probably dislike the Finance department but should not be surprised when they are offered a job in Finance (or Management). After all, Finance is really all about optimizing the business…

    Posted January 19, 2012 at 3:44 pm | Permalink
  3. Ben,

    Will do. Many of these things have been discussed in posts already up, but the graphic design one, in particular, I haven’t dedicated time to. It’s on the agenda now.

    In a nutshell, I have been most successful with using metaphors in design. Showing pictures of the metaphor while describing its applicability to the real concept. An example might be when I was trying to show someone how prioritizing lower, but more valuable traffic channels over high-volume, low value channels, I showed a cross section of 2 piggy banks: one was small and packed with cash. The other was enormous and had a buffalo nickel in it.

    David Armano also has great visuals that make things click. Funny enough, a lot of the inspiration comes from ridiculous or funny visualizations, too. Infographics, especially if you hate them, can be a great resource because you can usually describe the parts you like and the parts you don’t.

    For simple charting, if you can use metaphor instead of lines, that’s great. If you can’t I have a few “death mistakes” in charting, like NEVER connect two points on a chart if there is not a physical connection between two things or data in-between. Like when people do a line chart of revenue by channel.

    The most important concepts I’ve been working on in design are typography (often WAY more powerful than images, and I suck at it), directionality (make sure people digest things in the right order), and color (never make something unimportant look important, use color to make things look good, bad, or even “dead” or “pukey” – I’m looking to cause a physical reaction to information).

    Type should be used as a design element, not as content. Your content is what you are there for. Get rid of non-design type. Stop using PPT and start using Keynote. In Keynote, never resize any fonts that are pre-built into the template. If you are having to resize to fit more text on a slide, someone in your audience will attempt suicide. If you have more than 3 bullets, expect the audience to have a full clip.

    I’ll do something more concrete when I have the time a post like that deserves. But get good books on presenting (the one mentioned in the post is Slideology). Go get a coffee at Barnes and Noble and read design books in the aisle before they go out of business. The biggest lesson I’ve learned about design is that designers rarely just sit down and create something amazing. They search for inspiration and build on it. When I started doing the same thing, I started getting it a lot faster.

    Posted January 19, 2012 at 4:42 pm | Permalink
  4. First, I love the way you write. It’s so much fun.
    Second, awesome article – as usual.
    Last – I too am about as good at graphic design as Helen Keller :)

    Posted January 20, 2012 at 8:18 am | Permalink
  5. Nancy Koons

    Great post, Evan. I read this quickly the day you published, and today took the time to re-read it thoroughly. Putting the focus on yourself as an analyst, your audience, and what you are trying to accomplish (plus how you present those ideas visually) really is the core of what we do, but it is so easy to get toooo caught up in the tools and reports. Thanks for the refreshing reminder!

    Posted January 21, 2012 at 4:42 pm | Permalink
  6. This is totally going to come across as un-genuine (if that’s even a word), and I don’t mean it to… but I’m forever inspired by your brilliance. Like Ben, I’ve learned many of these lessons already but still found many good points in here to look for in future projects. I appreciate you taking the time to write this.

    Posted January 22, 2012 at 4:23 pm | Permalink
  7. Interesting post. I am agreeable with the above points. I am into web analytics myself and must say it is an art getting it right to yield desired ROI.

    Posted January 24, 2012 at 1:00 pm | Permalink
  8. anon

    Hi guys, have you ever tried to print a webpage of your website? I was just triying to print this page to a PDF file to read it later, but it seems that there is some problem with your css while printing..

    Posted January 25, 2012 at 11:41 am | Permalink