You see, Molly Meadows and I won a blue ribbon at the science fair in the sixth grade.

Our project, Soggy Cereal: A Breakfast Mess, started the way all good science does – with market research and a hypothesis.

The our methods and the execution during testing were questionable (we definitely cheated and logged some additional results after we decided to give up and explore the college campus where her mom was teaching). The general methodology looked like this: Molly and I dumped 3/4 of a cup of milk over Rice Chex, Corn Chex, Wheat Chex, and then Crispix. We set a timer and poked the cereal every 30 seconds to see how long it took for the cereal to become saturated with 2% milk. After 5 rounds of tests (several fraudulent), we entered our results into Excel and began writing up our conclusions.

The charts in our final presentation looked something like this:

3D Charts and Windows 95 Default Colors are the worst.

Luckily my graphing skills have come a long way, but so has my understanding of empirical evidence. It’s interesting to watch it informing my work now. Startup life has a decidedly empirical backbone.

• You have an idea.
• You research the market and associated risks.
• You build something.
• You test.
• You iterate.
• You rinse and repeat.

It’s a really cyclical process, and somehow the results can still manage to surprise you.

I wasn’t born a data loving nerd; I started college as a double major in English and biology. That plan changed within three weeks of starting classes.

My parents pushed hard for the biology track. They felt that in order to get a job when I graduated, I needed to major in something that would leave me with quantifiable skills. Then I got back my first intro level bio test. I informed them in short order that I was going to be a theatre major.

Despite taking only two undergraduate math courses (one Pass/Fail, to avoid having to put much time in), within three weeks of starting graduate school I decided to add quantitative methodology as my secondary field of study. And amid theory and language classes, I wound up taking 4 semesters of statistics.

So, how did I come to love statistics after thoroughly giving up on math?

The answer is simple: I love stories. At first glance, math seems intangible. Numbers are abstract until you can find a way to get them to mean something. Once you move the numbers from a data set into a chart or into a table, they become more than numerals – they show patterns of behavior, give insight about how to solve problems, and even present new and exciting mysteries to solve. Statistical analysis makes numbers into narratives.

At school, I was soon able to look at data and sleuth out answers that applied to real world problems. I could look at a data set, clean it, scatter plot it, slice and dice it, and then tell you (with reasonable statistical certainty) that, for example, there is a correlation between successful social movement organizations and malleable organizational structures or diverse funding sources. Or that self-identified cat people are more likely than dog people to skew urban-dwelling, female, and to name George Harrison as their favorite Beatle (guilty). I’ve got the numbers to back it all up.

Then digital advertising found me. The nature of the questions changed, but the end result didn’t. The questions were less about organizational structure or the likelihood of violent conflicts, and more about an advertiser’s return on investment. The questions became:

• What’s working?
• What could be better?
• Could I be spending my advertising dollars more effectively?
• How can I best present this data to my clients?

And, for me, how can I take what I have learned, and use it to inform future decisions?

The beauty of online advertising is that between analytics solutions and highly trackable media, we can tell what’s working and what’s not. Then we adjust accordingly. The end result: campaigns that just work better for clients. Campaigns that meet real needs and tell better stories. And in a world where more effective use of ad dollars not only means more revenue but more satisfaction for the end customer, everyone wins. That’s why I love stats; they help us shape our stories into what we want them to be.

With the proliferation of both national and local brands using social media to engage with their consumer bases, Get Satisfaction recently created an infographic citing data from Razorfish, Econsultancy and SocialMediaToday that details why customers connect with brands online and the effects these connections can have.

Not surprisingly, nearly 70% of people who connect to businesses and brands on Facebook and Twitter cite being either a current customer of the business or following brands to gain access to special offers and deals. An additional 20% of respondents pointed to the interesting or entertaining content posted on Facebook and Twitter as being the primary reason they follow brands.

Furthermore, between 60 and 70% of people who connect to brands online say that they consider these brands when they are in the market for new products, often wind up buying products or services from these brands AND recommend the brands the connect with to their friends and family. A staggering 97% of respondents said that their online experiences affected later purchases from a specific brand or business.

The takeaway?

(1) Social media is a powerful customer loyalty tool. When your customers are engaging with you online, they are saying that they not only endorse your work or product, but that they actively want to hear more from you.

(2) You should absolutely be leveraging your current client base to help build your social media presence- and you should make sure you are rewarding them with a special deal or offer every now and again.

(3) You don’t have to be afraid to crack a joke or offer up some industry-related tips & information- by connecting to you, they are showing interest in what you have to say.

(4) Though not always immediately quantifiable in terms of dollars, social media can certainly help with this ambiguous thing that is customer loyalty. In short, it keeps ‘em coming back.

Both Google and Bing have begun including social results in their search algorithms. Things you and your extended social networks have “liked” or shared are more likely to show up in your searches. Input from other users on your social graph will continue to affect the search results you are presented with, just as your own preferences will continue to inform the information you are presented with.

If I were to search something like “twitter and google”, I’d find that not only is there information about the partnership between the two, but also a Tweet that my best friend posted earlier this month.

See! My bestie's tweet on Google.

Here’s how it started. Within a few hours of Bing announcing in 2009 that they were going to incorporate real time Twitter feeds into their search results, Google announced their own non-exclusive search deal with Twitter. Google has since proceeded to include a real-time feed of tweets in their results pages, allowing for up to the second crowd sourced information about what is happening enabled by Twitters 140 character microblogging platform.

Google has, in addition to including Twitter results in their “real time search results”, expanded their algorithm and results page to include the Twitter feeds and Flickr pages of a searcher’s friends. Social search results are then blended among the other results returned by their algorithms. The thought is simple people on the same social graph are likely to share commonalities- be they geographic, interests or matters of taste.

Furthermore, in December of 2010, Bing announced that they would be including Facebook “likes” as a part of their own search algorithm used to rank search results. This was important for several reasons, first because it signaled a major move toward more personally tailored search results. The thought being that were searching for “Metric” and you have liked that band on Facebook and shared a link to one of their Youtube videos, chances are that you are looking for the band and not the measurement system.

Great band. Great band.

And secondly, the Bing-Facebook partnership is of note because this was the first time Facebook allowed a search engine a greater look into the tremendous piece of the private web that they own. Facebook and Google have been notoriously at arms with each other because of Facebook’s closed network- that doesn’t allow Google to index the tremendous amount of private data and internal pages on the site. This has been a major shift in strategy, and might also be signaling a Facebook shift to move outside of their walled off web property.

Google has also added Hotpot to their arsenal. The idea is simple, rate places you go, shop, eat, etc., by what you like and what you don’t like, then Google can suggest other things that you will like. Hotpot has also been incorporated into local searches, and has begun suggesting local businesses based on business you and your friends have rated.

So, when I search “brunch places lower east side”, chances are that my results will differ from your. It’s a matter of taste, and the integration of social search into Google’s algorithms.

Reposted from my Cast page:

In mid-November of last year, Google unveiled their newest addition to Google Places, Hotpot. The aim was to provide a service that compliments both Google search and Google Places by giving users local suggestions based on their personal tastes and preferences. In short, Google is aiming to update their search results to include personalized recommendations for local businesses.

Hotpot is relatively similar to several other local places and check in services, and with the addition of both an Android and iPhone app, Hotpot has been compared by other bloggers to Yelp, Foursquare, Bizzy and Facebook Places.

The idea is this, based on your location and a series of local places/businesses that you and your friends have rated, Google will be able to serve up a list of other spots of interest. Because of their tailored nature, these results should be more valuable than the previously general results a consumer would otherwise be receiving.

Having entered several reviews, and rated a number of businesses, my now personalized results are alright at best and got better only after fiddling with the site for well over an hour and a half.

The Issues:

1.)    After connecting with several of my friends on Hotpot, I found out that I have wildly different taste in bars and restaurants than they do.

Some of this is due to the vegetarianism I have enjoyed thoroughly for most of the last 15 years, and that my friends may live in different parts of the city than I do. The bars and restaurants I go to most often are closer to my neighborhood (Williamsburg, Brooklyn) than theirs (Battery Park, Park Slope, the Upper East Side and the Financial District)- even though we all live in the same city and within miles of each other.

2.)    The places of interest Hotpot suggests are predominantly tourist attractions.

These are places I only visit when entertaining out of town guests. I have no interest in visiting them regularly. When I click the not interested button, others pop up to replace them.

3.)    Hotpot doesn’t discern between the loyalty businesses it’s recommending and frequency businesses.

No matter how many times I have hit that same not interested button, Hotpot continues to show me auto service stations and hair salons. There doesn’t seem to be a learning algorithm that takes into account the fact that the reason I am hitting not interested is probably because I already have a salon and mechanic in my arsenal of favorite local businesses, or worse, I don’t have a car (or hair).

Additional things Google might want to take into account:

1.)    Frequency of use.

I am inevitably going to have more visits to various restaurants or bars than to the MoMA or my salon. I can eat out a number of times a week, I really only need a haircut every couple of months or so.

It would make sense to show me more of the businesses I can consume more of. Likewise, I might disproportionately want to rate these businesses as well.

2.)    The real difference between a loyalty business and a frequency business.

Similarly, I may have a favorite sushi restaurant, my loyalty isn’t strained when I try a new sushi restaurant, new clothing store or a new coffee shop. However, most people are relatively loyal to their hairdressers, auto mechanics, gyms and dry cleaners.

When showing me recommendations, Hotpot would be wise to take into account what kinds of businesses with which people are more agnostic about their loyalties.

3.)    The strength of a recommendation.

When people say they trust the recommendations of their friends, it’s assuming that friends will actually have the same taste in food, shopping and places they’ll want to spend time. The truth of the matter is that my best friend is a whiskey and bacon lover, and my relationship with both whiskey and bacon is tenuous at best.

What could easily fix this flaw is to show greater preference to personal ratings and the reasons behind them, regardless of that steakhouse my friends like.

And for that matter- yes, I may like the Apple Store, but it’s probably because I am a gadget geek. Because of this I am more likely to like a store like Digital Fix than the Chelsea Market. Correlation does not equal causality, and proximity does not equal similarity.

Things on Hotpot that I love:

The Saved Places List.

Period.

I can access it while mobile from my Droid with the newest update of the Google Maps application, or can log into my Hotpot account if I want a reminder of what that new restaurant I wanted to check out is called. No good will come of this particular feature (in the best and most delicious way possible).

It is also no secret that I am often swayed to do certain things because of simple conveniences. I am intensely self-interested in that regard. So, it’s no surprise that the yoga studio I go to (www.namastewilliamsburg.com) is but a few blocks from my humble Williamsburg apartment. It’s close enough that I can stumble to the studio half asleep in my pajamas (which just so happen to look a lot like yoga clothes) first thing in the morning. They’re affordable, they’re close, and their classes are great. I couldn’t ask for much more.

What is a surprise is that Namaste was not the first NYC yoga studio I tried to go to. In fact, in an attempt to be loyal and go to one of my clients for my yoga fix, I was totally game to travel to the West Village (15+ minutes away) until I discovered this:

They charge EXTRA money for allowing me the convenience of scheduling and paying for my classes online. It amounts to an extra 2 dollars per class. Thanks, but no thanks, dude. I am saving you time and staffing costs by taking the initiative to schedule online. Where I understand that it may be more expensive to build, operate and maintain your website because of your online scheduler, this is also ridiculous. I recognize that it is only two dollars, but given the amount of yoga that I do, that $2 will add up pretty quickly.

So, I went elsewhere. I go to a studio that keeps their notes in a rolodex (Bless their hearts!), and doesn’t require me to schedule my classes ahead of time at all. Truthfully, I couldn’t be happier with my decision.