Top takeaways from the Predictive Analytics Innovation Summit

I work in Digital Analytics. We are a relatively new field (or at least, a relatively new application of analytics) and one where I sometimes feel we don’t leverage enough of what’s been done before. Rather than re-invent the wheel, I headed off to the Predictive Analytics Innovation Summit in San Diego, to learn from a wider variety of industries, and see what they were doing with predictive analytics.

Here were a few of my key takeaways:

It’s not about the tools

I honestly thought this was just an affliction of the digital analytics industry, but it turns out, this is an issue in all area of analytics.

It’s not about the tools, it’s about two things –

1. The talent. Talent matters more than the tool. The “best tool” is one that you can hire great people to use. People are what will make analytics successful, not the tool.

2. Your needs. The “best tool” is the one that suits your business needs – not the most popular, the most expensive, or even the one with the most features. (If you don’t need those features, what’s the point?)

It is about strategy and culture …

Success with analytics isn’t driven by the tools you buy, the process you implement, or the technology you have – it’s about culture.

The creation and hoarding of data is not what will make you successful with analytics. What matters is the consumption of analytics by the organization, and allowing data to challenge beliefs and theories. It’s not about big data – it’s about using it to make big decisions. 

Case in point? Improving the efficiency of a predictive model is only worth a few percentage points in your model’s accuracy. But nailing your fundamental strategy? That’s where success comes from.

… and about communication …

 Analytics is also destined to fail if it’s not communicated well. This is often where hiring for analytics fails: too often, analysts are hired who can’t communicate with others in the organization. If anyone walks out of a meeting after an analytics presentation and doesn’t know the outcome or the next steps, that’s not their fault – it’s our fault.

Businesses today are in a difficult situation – we have too much data, yet too little knowledge. Analytics is critical to helping to understand what’s important, draw insight out of volumes of data, but without the right people and the right communication, you’ll never see that value.

Conference time!

It seems that Feb-Mar was a busy #measure conference “season”, and another is upon us! As harrying as it is to try to keep travel schedules straight, I must admit I’m really excited to attend and be involved in these  upcoming events.

 

 

Keystone Solutions Speaker Series
“Providing Relevant Experiences without Stalking”
(Panelist)
September 26, 2011
Austin, TX
Details

This one day event looks to be a blast. Not only it is hosted by the lovely Keystone Solutions boys and girls, but there are some great speakers lined up: Evan LaPointe, Emer Kirrane, John Lovett, Nicholas Einstein (as Evan puts it, “C’mon, the dude’s name is Einstein!”) and Lee Isensee. I’ll join John, Nicholas and Lee for the panel (I’m not worthy! I’m not worthy!) and am really looking forward to it. It’s a low-cost event, so sign up now!

eMetrics NYC 2011

(“Analytics Career Development” panel)
October 21, 2011
New York City
Details

This panel brings together folks from the small client, larger client and agency/consultancy perspective to talk about career development in analytics. What does a typical career path look like? What should you do to grow and develop, and find your next great opportunity? This session is all about YOU and your future!

 

 

Accelerate 2011
November 18, 2011
San Francisco
Details

Web Analytics Demystified‘s one-day FREE (yes, FREE!) event in San Francisco sold out within one day! But don’t worry, wait list places are still available, so go sign up! Some amazing speakers have been announced already, and there’s an opportunity to present in the “Ten Tips in Twenty Minutes” segment and win a $500 Best Buy giftcard! (Now, hmmmm. What shall I fill five minutes with?)

What are you waiting for? Education is just a step away!

 

Whose responsibility is online privacy?

Kissmetrics and a variety of its clients have been center stage in the news lately for tracking unique visitor behaviour, despite a user clearing their cookies. Shortly after the story broke, a number of high profile clients removed Kissmetrics tracking, arguably “throwing them under the bus” in the process. Now, Kissmetrics and more than twenty of its customers are facing a class action lawsuit, claiming the tracking violates privacy laws. However, there was  a similar lawsuit in 2009 over the use of “zombie cookies”, with some of the same businesses named as defendants.

This got me thinking, and into a rather lengthy debate/rant/conversation with fellow industry member Lee Isensee, which helped to shape (and refine somewhat!) a few thoughts around the responsibilities of the organisation tracking vs. the vendor providing tracking capabilities. While I find myself defensive of vendors and organisations that are being respectful of customers privacy, in line with the WAA Code of Ethics, the real question is:

Whose responsibility is it to protect consumer privacy – the business using the tracking, or the vendor providing a solution or product?

I can’t help but think – if you, as a company:

  • Choose a method of tracking that (many argue) violates users’ privacy and wishes
  • Don’t disclose the level of detail being collected, or how it will be used
  • Face legal action as a result of that tracking, and settle by agreeing not to use that technology again
  • Later, face accusations of similar tracking (similarly intentioned, though the mechanics perhaps differ)
  • But sever ties with the vendor, essentially blaming them, while claiming your company takes user privacy seriously

What conclusion is there to draw from that? Does it suggest that you, as a business, want to do that kind of tracking, and seek out vendors who provide those capabilities? (It’s a little hard to argue the “but we didn’t know” defense if you’ve faced legal action for this type of thing before.)

If that’s the case (and I understand this is a little difficult in the current climate) why not stand by this kind of tracking, disclose the approach and method, and explain the consumer benefits of it? Why claim to be privacy conscious and blame the vendor when your company has a major privacy backlash. You’ve previously chosen to engage in this kind of tracking (and faced the repercussions!) before? What leads you to do so again?

So if a business is inclined to this kind of tracking, what is the responsibility of the vendor providing it? Do they own a customer’s implentation (post initial engagement) or chosen use of the data? Do they owe a duty to the customers of their clients? What legal duty do they owe? Do they owe a duty to allow opt-out? Or is that in the hands of the company doing the tracking? What ethical duty do we impose? (And how far does that go? To the vendors that support the vendor? Ah, forget it. I’m hearing an Adam Carolla “slippery slope” rant starting as it is.)

I’d argue there’s one level of responsibility, that falls squarely to the company itself. A business decides what kind of tracking to do, and which vendors to use. They owe a duty to their customers. If a vendor is found to use “unsavoury” practices, actively recommending those practices in collusion with the business and disregards industry accepted practices, isn’t it the responsibility of the business to have thoroughly evaluated the vendor?

Something along the lines of: we don’t sue gun companies for homicides. The analytics vendor sells the gun, the implementation is the bullet, the business is the person holding the gun … who ultimately made the choice to shoot the customer?¹

Am I way off base? Where do you think this responsibility lies?

 

¹  I can’t take the credit for all of this. Thanks Lee for boiling it down to a simple analogy.

Analyzing the Impact of the Digital Fold

[Originally published on the ClickTale blog]

In the traditional world, we talk about the importance of being “above the fold”: appearing in the top half of the front page of a newspaper. However, on the web the picture is a little murkier. Website visitors will use different screen resolutions, browsers, window sizes and toolbars, essentially leading to a different “fold” line for every user.

Add in the proliferation of devices (desktop, laptop, tablet, smartphone) and the challenges are further compounded. So is there even the same impact of content being above or below the fold for online users as there is in the traditional world? Might this impact vary by user, by site, or by page?

Staying Above or Scrolling Below the Fold?

On pages such as a home page, the location of content above or below the fold may have a greater impact. After all, when a visitor arrives to a site, they need to figure out what content or products to dive deeper into. In this case, products or content areas highlighted in the top area of the page may receive higher engagement, simply due to the higher number of “eyeballs” on it during an evaluation phase.

However, the same may not hold true for deeper pages within the site, or for all-in-one landing pages. On a product detail page, where reviewing the content on the page may be crucial for making a decision to proceed to the next step, the click-through rate for a call to action at the bottom of the page could potentially be higher than a call to action at the top of the page.

Impact of Your Calls to Action

The impact of the fold may also depend on the call to action that you are measuring. For example, the ad click through rate may be affected differently by placement above or below the fold than lead submission or an Add To Cart call to action.

How To Analyze Actual Behavior

Design can play a huge role. In some cases, site design may make it clear to the user that there is content below the fold, and encourage content consumption lower on the page.

We can’t just assume the fold affects our site, or that it affects all pages in the same way. We’ll want to start by analyzing actual behavior.

1. Examine your site’s pages (or types of pages) separately. The behavior on your home page may not be the same as your landing pages or product pages. Start with your most important pages and go from there.

2. Use a traditional web analytics tool to give you an idea of the device and screen or browser size your visitors typically use, to start to understand where they see the fold on their machine. However, as you analyze, keep in mind there is no true fold – it is different for every user based on their settings.

To make the data feel more real, change your own computer settings to match your typical visitor, and encourage your creative or design team to do the same. Browse your site using these settings and you’ll get a better idea of what different visitors see and where on their screen it is located.

3. Ensure you are tracking individual calls to action on your pages in enough detail so you can understand (for example) above vs. below the fold performance. While many web analytics solutions will allow you to see if visitors moved on to the next step, if you have two calls to action on the page that link to the same next step, one above the fold and one below, you’ll want to be sure your analytics tool allows you track them separately.

4. Use an In-page analytics tool to understand interaction within your pages. While understanding click-through rate of your calls to action above and below the fold can help, that doesn’t necessarily tell you how many users actually saw the call to action.

5. Take time to segment this information. A good place to start would be the by the different screen and browser resolutions you have already examined. Try bucketing different settings, to analyze a group of visitors.

However, another consideration may be landing page. A visitor who has just landed on the page you are analyzing may be less apt to engage with content below the fold than one who has pathed to the page looking for specific content or products, and is looking to dive into detail about these.

Users looking at different products may also show different behavior. For example, a $5.99 purchase may require less engagement with product details and result in less below the fold engagement than a product that is $599.

6. Start testing. Once you have insights from these sources, you can begin to test the impact of changing them. What if you remove some of the content and make the product detail page shorter? Or move your call to action above or below the fold, or test having one above and one below? What about the left vs. right hand side of the page?

Conducting A/B or Multivariate tests of your layout, and tracking the behavior of these separately, can give you much more insight than pure analysis, because you can see the impact of actually changing things.

Overcoming the Complexity of the Digital Fold

There is definitely a complexity to be managed in analyzing the digital “fold”, but there are also great solutions out there to help us better understand user behavior within the page, and to optimize it for a better user experience and business results.

 

Web Analytics Association Boston Symposium

Curious about the Web Analytics Association Boston Symposium that just took place on Monday 5/23? Well look no further for a little fun with Twitter analysis and an overview.

Let’s start with the data …

Overview:

  • 159 unique tweeters used the #WAABos hashtag between 5/22/11 9.24PM and 5/24/11 1.08PM
  • There were a total of 911 tweets to the #WAABos hashtag in this time
  • During the Symposium itself (Monday 5/23 from 1PM through 5.30PM) there were 793 tweets from 125 unique tweeters.
  • This translates to 176 tweets per hour, or 2.9 tweets per minute!
  • 87% of the tweets to the #WAABos hashtag were during the Symposium time (Monday 5/23 from 1PM through 5.30PM)
  • 59% of the tweets to the #WAABos hashtag were Retweets and another 18% of tweets contained mention of another Twitter user. (My, we’re a social bunch!)

Top 10 hashtag contributors:

@OMLee 17% of hashtag tweets
@michelehinojosa 16%
@ashkalei 11%
@RudiShumpert 4%
@kdpaine 3%
@CaseyChesh 3%
@jc1 3%
@Exxx 2%
@KeithBurtis 2%
@johnlovett 2%

Top tweet content:

[Ten points if you can find the word “Pirate” in there.]

Tweet locations:

From @ashkalei:  Map of where #WAABos tweets were coming from: http://bit.ly/mlgyry

Top Take Homes:

Web Intelligence
Suresh Vittal, Forrester

  • Customers are no longer linear, or staying in neat “swim lanes”. We have entered the “splinternet”, where users can connect via multiple devices, and we start bringing that data together for a more comprehensive view of our customers.
  • We need to move from web analytics to all-encompassing web intelligence.
  • Web analytics platforms are perfectly positioned to evolve into web intelligence platforms. Almost 90% of businesses are using or piloting a web analytics platform, and many use more than one. Now, more traditional online channels (search, display, email) are regularly integrated into web analytics solutions, and emerging channels (social, mobile, apps, video) are starting to be integrated.
  • Merging offline, traditional web and emerging channels will give us a  comprehensive view of our customers, and pave the way for web intelligence. (And yes, it’s complicated!)
  • Be guided by a roadmap, and be sure to consider process and the personnel and skills you’ll need, in addition to the technology. Web Analysts alone will not be enough.

Mobile panel

Raj Aggarwal, CEO, Localytics, Justin Cutroni, Director, Cardinal Path, June Dershewitz, Director of Web Analytics and Customer Insight, Apollo Group, Mihael Mikek, CEO, Celtra

  • Mobile is currently fragmented – apps, different operating systems, web. In a year or two, we won’t even be talking about “mobile” – everything will be connected.
  • Your users don’t differentiate between a mobile and non-mobile experience, so you need to integrate your digital strategies.
  • The Three A’s of Mobile: Awareness, Activation, Activity (Apollo Group, June Dershwitz)
  • But these must also be tied to your overall business strategy.
  • Next problem for mobile to deal with: cannibalisation. Are you stealing from other channels or is this new revenue?
  • Mobile apps or mobile web? Right now, mobile apps are superior because you can integrate with other features of the phone (e.g. address book, etc.) However, HTML5 will rebalance that and it is likely that browser based apps will take off vs. OS-specific applications.
  • Difficulty for analysts is understanding behaviour from mobile to web and other channels, as mobile data typically lives in a silo. Crucial for us to start understanding behaviour of users across channels.
  • We can learn lessons from the web, to speed up the learning curve.

Social media panel

John Lovett, member of WAA Board of Directors and Senior Partner, Web Analytics Demystified, Katie Paine, CEO, KD Paine & Partners, Sean Power, Founder, Author, and Consultant, Watching Websites

  • Social can be many things to many people or organisations. This requires the need for custom metrics and integrations.
  • However, the web analytics problem of silos is repeating itself with social. There is isolated use of social media in the depths, but not across the enterprise. (John Lovett)
  • Great debate between Sean Power and John Lovett: Sean argued social media does not scale – you can’t respond to everyone without hiring people to respond one-on-one. John argued companies like Dell are tackling this by teaching their existing employees how to respond. Sean tested this by tweeting Dell while on stage at the panel, to see how quickly they respond. (19 minutes, if you’re curious.)
  • Do you know what Pirate Metrics are? AARRR!  Acquisition, Activation, Retention, Referral, Revenue: http://slidesha.re/yO8Ml
  • Do we need social media standards?
    • Sean Power: “I don’t think businesses give a s*** about standards.” They care about making money and will do whatever they want.
    • John Lovett: We need at least some standards – definitions of basic, common metrics, even if different tools calculate them differently.
    • Katie Paine: We need standards so we’re not confusing others.
  • What about sentiment analysis? Sentiment analysis is like web analytics – you need the best people, not the best tools. (John Lovett)
  • Need context in social media. A small fly looks terrifying through a magnifying glass – which is what sentiment analysis can do. It’s important not just whether customers are saying something negative, but whether they are more negative about you than your competitors.

Tom Davenport: The New Quantitative Era – Creating Successful Business Change with Analytics

  • Analytics involves moving from descriptive analytics (the “what”) to predictive and prescriptive analytics (the “so what”)
  • In its most basic form, analytics is about making decisions.
  • Using data to make decisions, however, requires mastering analytics, culture and more. It’s no longer sufficient to just be good at one.
  • Become a student of error. Reviewing your mistakes can lead to better decision making.
  • To become successful at analytics, you need to work closely with IT, business decision makers and outside ecosystem members.
  • If you want to make decisions better, it’s not about the math, it’s about the relationships the analyst builds with decision makers.
  • Analytics and the work done should tie to decision. When an analyst receives a request, the first question should be What decision will you make with this data?
  • Skills needed to be a good analyst:
    • Tell a story with data
    • Stand firm when necessary
    • Help from the decision
    • Don’t just identify a problem, fix it
  • The analytics industry has a historical opportunity right now to transform our industries and functions!

Entrepreneurs Panel

David Cancel, CEO and Founder; Performable, Matt Cutler, CEO and Founder, Kibits, Eric Hansen, CEO and Founder, SiteSpect, Jonathan Mendez, CEO,Yieldbot, Dennis Mortensen, CEO and Founder, VisualRevenue

  • Let the market tell you what is right.
  • Everything I’ve done is based on solving customer pain. Can I give you an hour of your day back? (David Cancel)
  • You want a reaction to your idea. “I love your idea”, “I hate your idea”. “Cool” or apathy is not a good thing.
  • Commonly heard: “The last thing I need is another damn dashboard.” What they want is a red phone they can shout questions into.
  • Great companies are bought, not sold. The minute you raise your hand, your value goes down.
  • Charge immediately. From day one. The kind of feedback you get is very different the moment you ask for a dollars. (David Cancel.)

Any other insights that you heard that I missed? Add ’em in the comments!

Want to have at the raw data yourself? This is the archive I used: WAABosTweets052411at0113PM

 

 

A Penny For Your Thoughts on Influence?

[Originally published on the Measure Mob blog]

I was disappointed to not be able to join Keith, Jason and Olivier on the first two Measure “Mobcasts” – those darn international flights sure do get in the way! However, that doesn’t mean I won’t take the opportunity to put in my $0.22. (It would be my $0.02, but I was in Australia, and their dollar is worth more, so I figure I’ve got the exchange rate working for me.)

I wanted to add what are just a few (minor) “parting thoughts” after having an opportunity to hear the guys so deftly discuss their opinions on measuring social media influence. So here are mine …

1. Measuring Social Media “Influence” is Necessary

The reality is, while companies (and even individuals) would love to have an opportunity to engage with every voice that reaches out to them via social media, that kind of engagement isn’t necessarily scalable or realistic, and even if it was, there would still be a need to prioritize the order in which a company reaches out to people.

While that doesn’t mean, to Jason’s point, that anyone should be ignored because their Twitalyzer scores aren’t high enough, some kind of measure of where to start is realistically necessary, especially for companies with a large number of social interactions.

2. Social Media Influence Measurement Isn’t Perfect

I feel like part of where we get hung up is in thinking that if a measure isn’t perfect, it’s not useful. I agree with the guys that a measure of “influence” should be considered in context of other data, and I also agree that it’s probably more realistic to call it “potential influence.” After all, you never know whether someone who is considered to have social media influence is actually going to influence behavior of fans/followers/friends.

As we love to say in the analytics industry – you can’t manage what you don’t measure. However, the unfortunate reality is that you can’t measure everything that you would like to manage. (And our attempts to do so often end with “KPI” Dashboards that show fifty metrics instead of the one thing that executives want, because that measure isn’t truly possible at this stage.)

How could we measure true influence? Maybe: Person A engages with Person B, or shares a positive experience with everyone. Person B then goes, “Huh – I hadn’t thought of going to Restaurant X for dinner” and heads on over. Well, sadly we don’t have that insight. (Yet, or maybe ever.)

So what do we have? We have measures that look at, in the example of Twitter, how many contacts someone has, how often their tweets are shared or responded to, as a proxy for influence. Does that mean that the person following will “monkey see monkey do?” No. Is the measure perfect? No. Is the measure useless? No. Understood for what it is, it can be helpful. Blown out of proportion, of course it’s not. However, I know one thing – digital measurement is a constantly evolving industry. We will get better at this. But that doesn’t mean that we can’t do something with our “first draft”.

Red Door Speaker Series: Digital Measurement

On Tuesday 4/12 (San Diego) and Thursday 4/14, Red Door Interactive held our regular Speaker Series. This time, the topic was one dear to my heart: digital measurement.

We had some fantastic speakers agree to be a part of our San Diego and Denver panels, and a huge thanks goes out to them:

San Diego panel:

Denver panel:

Topics

Our two panels discussed a wide range of issues related to digital measurement, including:

  • Their definition of “success” and focus of digital measurement practice in their organisation
  • Challenges and frustrations of digital measurement
  • Measurement and planning strategies for new channels (for example, mobile and social have been key of late)
  • How they cope with the feeling of “drowning in data”
  • The challenges of data silos and getting a more holistic view of the customer
  • How to sell a data-driven idea internally
  • Educating business users as to what the analytics team can provide
  • Where to focus additional investment in digital analytics. For example: Tools or people? (Hint: the resounding focus was on people!)
  • Online privacy concerns and the impact to digital measurement

Some great takeaways from @reddoor on Twitter:

A “Thank You”

Thank you to our great speakers for finding the time to come and contribute to the event, and a special thanks to Rudi Shumpert for traveling so far to be with us!

And while you’re at it, check out Rudi’s recap on the Keystone blog.

Pictures

Speaker Series San Diego

Speaker Series Panel, San Diego

Speaker Series Denver

Speaker Series Panel, Denver

More photos are available at the Red Door Flickr account

Omniture Summit 2011 on Twitter (Day 1)

So, because I’m a huge nerd (and I assumed others might be too) I thought folks might enjoy some information on #omtrsummit (aka the Adobe Omniture Summit 2011) on Twitter.

Half way through the opening session today (I’d say around 9.30AM Utah time) I started a hashtag archive using Twapper Keeper.

Some completely fun but not very actionable findings:

Approximately 17% of Summit Attendees tweeted: 441 unique usernames tweeted at least once, compared to 2600 attendees. (Note: I’m sort of assuming that if you didn’t tweet in the first day, you’re not likely to throughout the rest of Summit, but I’ll gladly check those findings on Friday!)

Top 10 Tweeters, in order of volume of total tweets:

dennisy
omtrsummit
michelehinojosa
RudiShumpert
johnrmatthews
EndressAnalytic
ad0815
bill_ingram
pvanhouten
c_sutter

Total 10 Tweeters, excluding retweets/via:

dennisy
omtrsummit
michelehinojosa
EndressAnalytic
kennovak
craig_burgess
spike96
pvanhouten
lorriegeek
theshammond

Oh yeah – and 1.4% of tweets on Day 1 included a reference to Charlie Sheen.

 

Want to grow the analytics workforce? Go out and get ’em

Yesterday, Red Door Interactive held an intern event. We have four paid internship openings for this summer, in four fields (marketing communications, email marketing, SEO and web/digital analytics.) The purpose of the event was two-fold:

1. Tell the interns a little bit about the different positions – What is web analytics? What is SEO? – and give them a chance to learn, ask questions, and figure out what they’re interested in.
2. Get to know the interns, to start the interview process.

When interns came to the event, they had previously applied and indicated which of the positions they thought they might be interested in. Not surprisingly, a large percentage were Marketing Communications – because who knows what web analytics or email marketing is if you don’t already work in the field?

What was amazing was how many spoke to me after the event or emailed to say, “I put down Marketing Communications, but now that I know more, Web Analytics sounds really interesting to me!” or furiously wrote down my web analytics book recommendations.

So here’s where I got to thinking. Those of us in the industry, especially those of us in a manager, director, VP level (etc), lament the lack of smart, qualified people, and how hard it is to hire. There are too many positions and not enough good people to fill them.

Well you know what? We need to do something about that.

So here’s what I am going to do. I am going to reach out to local universities and colleges, and see how I can start getting in front of students. I am going to tell them about our field, what we do, what the work is like, what a “day in the life” involves. Some students will shrug, but I guarantee that some students’ eyes will light up, just like they did yesterday.

We’re not going to encourage more smart young people into our field by hoping they stumble upon analytics. We’re not going to grow our industry by random chance. We need to go out and get them – and there’s no time like now.

Who’s in?

It’s not just about the race, it’s about how you train

This is one of those things that has stuck with me ever since I read it. It’s an account from a Colombian rider, Santiago Botero, in the Tour de France, about being overtaken by Lance Armstrong in a particularly tough mountain stage.

We all get so busy managing the day to day of our work, and the crises that come up. But the truth is, we need to find time for the long-term efforts , to think forward as to what initiatives we can start on now. It is tough to prioritise these when they won’t ease the present pain points, but they’re crucial to our development in the future.

It’s not just about the race, about that present moment. It’s about what you did to train, to prepare, to be ready for future challenges.

There I am all alone with my bike. I know of only two riders ahead of me as I near the end of the second climb on what most riders consider the third worst mountain stage in the Tour. I say ‘most riders’ because I do not fear mountains. After all, our country is nothing but mountains. I train year-round in the mountains. I am the national champion from a country that is nothing but mountains.

I trail only my teammate, Fernando Escartin, and a Swiss rider. Pantani, one of my rival climbers, and the Gringo Armstrong are in the Peleton about five minutes behind me. I am climbing on such a steep portion of the mountain that if I were to stop pedaling, I will fall backward. Even for a world class climber, this is a painful and slow process. I am in my upright position pedaling at a steady pace willing myself to finish this climb so I can conserve my energy for the final climb of the day.

The Kelme team leader radios to me that the Gringo has left the Peleton by himself and that they can no longer see him. I recall thinking “the Gringo cannot catch me by himself.”

A short while later, I hear the gears on another bicycle. Within seconds, the Gringo is next to me – riding in the seated position, smiling at me. He was only next to me for a few seconds and he said nothing – he only smiled and then proceeded up the mountain as if he were pedaling downhill.

For the next several minutes, I could only think of one thing – his smile. His smile told me everything. I kept thinking that surely he is in as much agony as me, perhaps he was standing and struggling up the mountain as I was and he only sat down to pass me and discourage me. He has to be playing games with me. Not possible.

The truth is that his smile said everything that his lips did not. His smile said to me, “I was training while you were sleeping, Santiago.” It also said, “I won this tour four months ago, while you were deciding what bike frame to use in the Tour. I trained harder than you did, Santiago. I don’t know if I am better than you, but I have outworked you and right now, you cannot do anything about it. Enjoy your ride, Santiago. See you in Paris.”