Hulu’s Lip Service

Recently I wrote about Hulu’s missed opportunity in not acting on the feedback they receive on the ads shown on the service.  In fairness to Hulu, I neglected to mention that they have a feature called Ad Swap, which in concept is innovative and compelling.    According to Hulu, Ad Swap “puts complete control in the hands of the user by enabling them to instantly swap out of an ad they are watching for one that is more relevant.”  In other words, they are giving the viewer control over the messaging that “hits them” by giving them the opportunity to choose it.

Sounds good, right?  If you’re forced to see an ad in order to view content, why not choose something you like?  You’d have to think that most people would take that offer.  But let’s look at the numbers, which shed some light on how well Hulu delivers on the promise.  Since its launch in October 2011 over 9 million ad substitutions have occurred on Hulu Ad Swap, according to MediaPost.  That figure seems significant but it’s almost nothing when compared to the massive number of ads shown on Hulu.  This week Forbes reported that in March 2012 alone 1.7 billion ads were shown on Hulu.  That’s 1.7 billion ads in one month.  Seven months have passed since Ad Swap launched.  So assuming Hulu averaged a billion ads a month during that time (and more in March), that would equal 7.7 billion total ads (6 months at 1 billion each + 1.7 billion in March).  It’s probably more.  But during that time only 9 million ad substitutions have taken place.  That’s a substitution rate of .12%.  

Do you really believe that 99.9% of the time Hulu viewers decided not to choose a more relevant ad?  Do people really want to see a bad one?  It seems that either Hulu is just giving lip service to “customer choice” or doing a terrible job of promoting their own feature designed to provide that choice.  And that’s the point of my post called Hulu’s choice.  If you’re going to ask for audience feedback, then do it well and act on it in ways that really engage your audience.

Besides this, I have no beef with Hulu.  Frankly I enjoy their service.  But when a company goes public with a new feature and then brags about its usage, they open themselves up to scrutiny.  And in this case, that scrutiny seems fair.

Hulu’s Choice

Today I went on Hulu.com to catch a few minutes of the latest Saturday Night Live episode.  Naturally, they showed me a (forgettable) commercial before I was able to view the show.  And as usual there was a feedback button that asked, “is this ad relevant to you?”  I’d seen that before on Hulu and at first it seemed nice that they asked for input but it did raise the expectation that somehow, over time, the commercials would be tailored to my specific tastes or interests.  But no such luck.

After watching some of the SNL show, I had to sit through another ad which was also forgettable, and in that context, a big annoyance.  Hulu and NBC are making money by forcing viewers to sit through commercials they don’t want to see.  But how is that any different from regular TV?  It unfortunately is not.

The web is supposed to be different somehow.  Interactive.  Engaging.  A feedback loop.  But too often online advertising is just a mirror of traditional advertising with bad execution.

The benefit of streaming video for the audience is time shifting…the ability to watch your show when you want.  And for that privilege, viewers are forced to sit through commercials they don’t want to see, even after explicitly saying they don’t like them.  At least with regular TV the expectation of relevant advertising is not raised.  You know what you’re in for:  a dumb monitor showing dumb untargeted commercials with no feedback mechanism.  And that’s kind of okay.  No expectations, dulled down senses and an occasional laugh.

If you’re going to give people the opportunity to provide feedback, really do it.  Give robust options. Commentary.  The ability to see what others think — including the votes of your friends.  Show statistics broken out for one’s state, county and city vs. the national average. A map that shows where people liked the ad the most.  An invitation to make a better commercial, with some tools to make that happen.  Sure, most people won’t participate in these things because they just want to view content.  But there are ways of making feedback engaging for people who are so inclined.  Certainly those people may be in the minority but some of them just might have hundreds of Facebook friends.  And others may spark an online discussion that yields 10X more visibility for a brand than a static, traditional commercial.

When the medium changes, so should the execution of the marketing.  Anything less is just, well, less.

Google’s Semantic Shift – and It’s More Than Semantics

Yesterday Google announced that it will increase its focus on semantic search results that better understand user intent and show people more answers on Google itself (in addition to linking people off to other sites).  As The Wall Street Journal reported, with this shift “people who search for ‘Lake Tahoe’ will see key attributes that the search engine knows about the lake, such as its location, altitude, average temperature or salt content.  In contrast, those who search for ‘Lake Tahoe’ today would get only links to the lake’s visitor bureau website.  What’s behind this?  Google wants to increase page views on Google.com and sell more ads.  They will also help users get answers more quickly…. and “helping the user” is a good business model.

The good news is for B2B marketers is that in our business we see lots of long, detailed search phrases, e.g. “fixed field sensors for background suppression” and complex queries are harder for Google to database, which means they’ll have to link off to other sites.

What are the implications for search engine optimization?  I’ll leave that for the SEO gurus to sort out but, long term, it would seem to pose some real challenges.  My observation right now is on the way Google has handled the announcement.

I don’t recall them so explicitly signaling, er, warning the market before about keeping more people on Google itself. In a masterful PR stroke, Google has become really transparent about the thinking that goes into their search algorithm and the huge amount of engineering talent they have put into organizing the world’s information for our collective benefit. (1000 man-years of labor and as many calculations as it took to send a rocket to the moon).

Check out the video below. It’s an internal Google staff meeting with their engineers discussing the minutia of algorithmic changes.  With PR like this, Google is trying to send a signal that it means no evil and that no specific sites are earmarked in its algorithmic decision making — but we all know that big algorithmic changes inevitably result in some collateral damage and major traffic drops for some sites.

 

I’m just waiting for Facebook to launch an improved search engine…then things will get really interesting.

Keep Ideas Flowing to Get a Culture of Innovation

In case you missed it, here’s a piece I recently wrote for Chief Business Marketer on creating a culture of innovation.  A little context:  I believe that making innovation part of a company’s DNA takes a combination of top-down and bottom-up approaches:  change driven by management and by employees. Keeping ideas flowing is critical to both, not only in the strategy but the execution.  This post addresses the flow of ideas…which is not just a matter of sending a note to the troops but an overall philosophy that should be pervasive in any company.  But as I mention below, sometimes a note to the troops can help.

————————–

Marketing is driven by ideas. Today, with marketing automation, social media and Big Data, the tools of our trade are changing faster than ever. What matters more than the tools, of course, is how we use them. And it’s your people, hopefully, who have great ideas. Do you find that 80% of the ideas come from 20% of your people? If so, the question for you as a marketing leader is how you tap the creative juices of the whole team, especially those who are less vocal.

Recently I sent a note to my team (about 40 people with diverse backgrounds in direct marketing, technology, data/analytics, operations, market research, copywriting, design and other functions) as part of a campaign to bring more ideas forward. Consider this a pre-fab letter you can adapt and use to stir up the troops—like Mad Libs but without the bathroom humor. I hope you find it useful.

Team,

Our business must continue to transform. And you have ideas. Like ______ , ________ and __________ (noteworthy initiatives suggested by your employees).   You may at times be reluctant to bring an idea forward, which prompts me to write this note. Some ideas will succeed and many will fail—but we’ll never know until we try. And focused experiments are how we will continue to innovate. So in order for (Your Company) to put more ideas into action I encourage you to share your thoughts, develop them or convince others to.

In my experience there are two overall ways of working:

1. Keep your ideas to yourself
2. Share your ideas, develop and execute them

Which one describes you?

If you keep ideas to yourself, there could be a couple reasons (but you should be aware of the implications):

1. You think your idea is not good enough or nothing will come from it—although certainly nothing will happen if you stay silent.

2. You want to develop it further before sharing. That’s an admirable intention and I’m all for ideas that are based and informed by facts, insights and data. Just be aware that long ideation and development cycles generally don’t work in our business. Keep this in mind if you feel a need to research, analyze and refine the idea and keep it close to the vest until you have imagined every permutation possible, considered every counter argument and obstacle conceivable and developed a bullet proof plan with operational focus and six sigma quality standards that Jack Welch would drool over.

Because by the time your perfect plan is ready, something has changed. ______ (a competitor) buys ______(a start-up).  ______ (a company in your market) launches ______ (a new product). Three mobile apps come on the market very similar to your concept. So, what’s a creative person like you to do? Put our company values into action…________, ______ and ______ (your company values).  So I encourage you to share your idea, get it out there. Talk to a colleague. Get opinions from others. Talk to some customers. Get to know their jobs and challenges deeply.

When you do, a few things will happen:

1. Others will build on your idea and make it better.

2. You’ll know if it’s currently being done (and done well)—and if so maybe you can help improve on what’s being done.

3. You’ll learn if, for some reason, there are other priorities. That would not be a bad reflection on you—you might just be ahead of your time or the company overall may simply have other things to tackle now. (In which case you can revisit it at another time).

4. If it doesn’t have the potential you envisioned or is not implementable—you’ll find out sooner rather than later, learn why and get some closure. If so, jump back on the horse and try again another time.

I’m not saying you should casually blurt out every thought that comes to you. Be smart and selective but share. In the ideation process—and in vetting new concepts—it’s also very helpful to speak to innovative young companies like vendors, developers and others that are paving new ways. Where can you find them?

1. _________ (blogs or other niche web sites in your industry)
2. Twitter
3. LinkedIn
4. Vendor webinars
5. Trade shows and conferences
6. Many other ways. Be creative in how you develop your creativity.

To build upon and refine your ideas, seek out the ideas of others, particularly when it comes to execution. Rally others behind your concept. To do so, it helps to simply walk down the hall, pick up the phone and introduce yourself to someone you’ve heard about (or someone you have not met yet). Or use _______ (your company intranet) or join _______ (your company’s networking group on LinkedIn). Be fearless. Miscalculations and failures will happen. Those are ok. We’ll just need to learn from them.

And remember, in a Culture of Sharing it is incumbent upon us all to keep an open mind, be constructive and give credit where credit is due. So I encourage you to get out of your comfort zone today, dust off that idea you have (or had), and share it with someone else. If nothing happens keep trying…with another person, with another idea or at another time. Our business must transform. And YOU might just have the next big idea. Or a seemingly small idea with big implications. So let’s get it out there and make it happen.

When Algorithms Rule the World

A friend of mine runs the development of risk management systems at a top Wall Street firm and recently I got some insights into his work.  Fascinating stuff:  he develops the models and systems that hedge the risk of the firm’s investments based on thousands of inputs and sophisticated business rules that manage their financial risk in real-time.  To him, a few million dollars is a rounding error, since the models hedge risk on billions of dollars a day.  Essentially, it’s a giant algorithm.  A model.

I believe algorithms will one day rule the world.  At least the business world.  Let’s take customer service.  It’s always nice when you get a little extra customer service at a store, a restaurant, a hotel.  Anywhere.  And these days, with the inter-linkages of customer touch points and CRM systems that track your preference for, say, fluffy vs. flat pillows, we’ll be seeing more algorithmic driven service.  That phrase sounds like a paradox because isn’t service, by definition, delivered by humans?  Well, in the age of marketing automation, with pre-defined rules that trigger auto messages it may not be the nice lady at the counter who sends you that lovely follow-up note.  And when it comes to the business of customer service, we should distinguish the delivery mechanism from the rules that govern it.  The business rules will increasingly be based on data models.

American Express is a good example.  They proactively call people when their algorithm notices a purchase that is a statistical outlier — an “unusual transaction” they say.  How good is their algorithm?  Someone I know always rounds up every restaurant charge on his Amex card to the nearest dollar. Instead of $56.81 restaurant tab (including the tip), he’ll add 19 cents to make the bill $57.00.  The rounding makes it easier for him to review charges on his Amex bill; if he sees a restaurant charge that includes anything with cents he knows it’s wrong.  Recently, in a rushed moment, he forgot his usual pattern and did not round up the bill.  The next day, American Express called him to inquire about an unusual transaction.  Their algorithm recognized his purchase was at a restaurant and that he did not round up the bill as usual. To him, it was a proactive security measure and really good customer service.  Behind the scenes, it was a really smart model.

Algorithms are everywhere.  When you call an 800# to inquire about a product and get extra speedy service, it may be due to a program that recognizes your phone number, associates it with a zip code and other data that predicts you are more likely to buy.  When you check out in a supermarket and receive coupons printed on your receipt, the offers are based on massive data sets of aggregated purchase history run through a model that predicts which coupons are most likely to generate future purchases.  Google has a $20 billion dollar business based on a model so elusive that an entire industry (search engine optimization) is built around figuring it out.  Amazon.com too, but that’s an obvious one. The notable thing about Amazon’s product recommendations is that they are determined in real time based on your online browsing patterns.  Click on a Kate Bush album and you’ll see other albums from Kate Bush or artists similar to her — not just in the emails you get afterwards, which has been around for a while, but during the same web site visit.

For airlines, pricing of airfares is based on an algorithm with a complex set of rules that optimize revenue yields for routes.   Prices – and price elasticity – vary by location, number of flights in the route, days before departure, booking levels, seasonality and of course competition.  Seems like the airlines need an algorithm to manage their algorithms.

These kinds of modeling applications have long been the purview of enterprises with deep resources:  in-house statisticians, data scientists, programmers and quality assurance — or funds to outsource the development of custom models.  But now it is it is easier to run models over large data sets than ever before.  There are many pre-built statistical models available for purchase that can solve for a wide variety of use cases out of the box.  And when custom models are required, programmers can create data analysis faster, with fewer lines of code, especially with programming languages like R which is designed specifically for data analysis.  It’s estimated that R is used by over 2 million analysts world wide.  That’s a lot of people developing code predicting your inclination for chunky peanut butter.  And now, in tech circles the term “big data” is almost a cliche at this point.

I believe that cloud computing, accessible statistical analysis and the age of apps will result in the democratization of models — an era when small businesses have easy access to algorithms that power their operations.

We’ll see a day when pizza shops predict with accuracy which toppings to include on their pies based on the hour of the day (and the optimal order of delivery drop-offs to reduce the costs of gas and speed up delivery times).  We’ll see a day when small online merchants optimize their merchandise and promotions in real-time just like Amazon.com. Powering the shift of algorithms down market will be service providers focused specifically on that goal.  Just as Constant Content did for email, Hubspot did for online advertising and social media, Google did for analytics and MarketTools did for online market research.  All of these companies brought high end solutions down market to mom and pop businesses.   The same will be done with algorithms.

And since 90% of all businesses are small companies, algorithms will rule the world.  Or will they?  Eventually, as algorithms become pervasive and we become an Economy of Algoritihms, the efficiency and effectiveness of models may result in their over use.  We may become overly reliant on the algorithms, because they will drive commerce, customer service and marketing.  But no algorithm will be able to detect the smile on a customer’s face or the furrowed brow of the annoyed prospect who waited too long in line.

It reminds me of the Star Trek episode where Captain Kirk and his crew are surrounded by enemy aliens who are “more advanced” than humans and have the crew out numbered.  With one look of the eye and without saying a word, Kirk gives a nuanced glance to his crew…phasers are shot, a fight ensues and you know the rest.  Victory for the Starship Enterprise.  When algorithms rule the world, ultimately, there will be a premium placed on wise human judgement.  Sure, the nice lady at the counter will know that you like fluffy pillows by looking in the hotel’s CRM system.  But it will be her, and her alone, who decides to offer you an extra cookie after that long flight.

Find Gold in the Data with Leading Indicators

Many restaurant owners can tell you, with decent accuracy, the revenue they expect in a given night based on how many customers are waiting for a table at any moment.   The number of people in line is a leading indicator of revenue.  And the restaurant owner who is that in touch with key metrics has more instinct than many large companies when it comes to their analytics.

When it comes to web data, organizations should think like the restaurant owner and look at the leading indicators.  If it’s repeat traffic you want, what are the factors that drive repeat traffic?  What kind of content, merchandising and promotions are likely to drive people back to the site over and over?  And if you don’t know the answer to that question, you should re-think the way you set up data in your analytics system.

When it comes to analytics, be careful what you measure…and what you don’t.  For example, if you are looking at repeat traffic make sure to factor out web site bounces because people who stumble upon your site accidentally and leave in 20 seconds are generally not the customers waiting in line at the restaurant.  They are people who come to your driveway only to make a U turn.  And you don’t care about the leading indicators for an unqualified audience.  Sure, you’ll want to dig deeper and find out who they are and why they left so quickly.  But when you factor out non-core traffic, the leading indicators that drive key metrics for the core traffic are more evident.  (Because there is less noise in the data).

Leading indicators help you probe more deeply than high level metrics.  They help isolate the things that can really drive your business and truly shed light on your web operations.   Just be careful when interpreting data to distinguish between loose connections and things that, with statistical significance, drive those connections.  That’s when art meets science.  And in the world of analytics, that’s where to find the gold.

The Ultimate App for Husbands

When shopping in a supermarket recently I overheard a man complaining about the complex rules established by his wife:   “Get dye-free detergent, no other kind.   Unless it’s on sale, then it’s ok to buy one with dye.   But only if it’s Tide or All.”  It made me think that this guy – and husbands like him – needs a database to store and manage all of these rules.  And it would sure help to have the database be part of a mobile app so husbands can easily put these rules into action when confronted with vexing decisions like whether to buy lunch sized vs. dinner sized napkins and whether to clean the kitchen table off with Fantastic or a Clorox wipe.  Husbands simply have to know the rules.

So drawing from my product development days I started chewing on how this app would it work and what it would do.  Here are the high level requirements I came up with so far:

• Voice recognition:   So spouses can speak into the husband’s mobile device and verbally input rules while he is watching Monday Night Football.
• Rules database:   It comprehensively stores thousands of rules established in the household.  This is the core feature.
• Semantic/fuzzy search:  Given the shoddy memory of husbands everywhere, natural language search helps him easily find “Clorox wipes” when all he can remember is “those slimy white things that are hard to pull out of the canister.”
• A massive, crowd sourced taxonomy:  Organizes the content of the rules which makes for fast and easy retrieval. Time is of the essence when you are putting the toilet paper on the holder and have to remember if it should roll over or under .
• Conditional logic:  Powering the app are a complex set of if/then rules for every possible contingency.  Like IF the kids had breakfast cereal yesterday, do not give them cereal today.  UNLESS there are fewer than 7 grams of sugar in the cereal and THEN it is ok.

Here are some other requirements:

• Social profiles:  What’s an app without a social component?  This feature would enable the husband to create a profile so he can connect with other creatures of his ilk.   Login would be with your Facebook credentials so you can immediately connect with 250 million other husbands who are equally inept. 
• Location aware functionality:
  • Location based push notifications – so when he is at CVS he can be prompted to pick up Qtips (entered in the system while he was oblivious and watching Monday Night Football)
  • Location sharing with randomized signaling – so any attempt to track his exact location can be foiled (unknowingly to the would-be tracker/spouse)
  • Four Square integration – so he can check in at local establishments approved by his spouse (admin rights with tiered access needed).   And if he goes there often enough he can save $5 at fine establishments like Chili’s.
  • User generated content:  So husbands around the world can contribute (and find) excuses for not following the rules.  A star system would enable the community to highlight excuses that are especially creative and plausible.  And users can search, sort and view excuses based on the  profile characteristics other husbands.  This way, a churchgoer in Alabama does not get advice from a swinger in New York.
  • Collaborative filtering/recommendations:   Excuses are proactively recommended for the husband even before he screws up.  This is based on an algorithm that determines the most effective excuses used by others who match the husband’s profile.  In phase II, there would be an integration with the male equivalent of Siri, which would speak the excuse for the husband if he is too busy watching Monday Night Football.
  • Machine learning:  unlike the husband the recommendations get smarter over time.

I am still working on the rest of the requirements and welcome your feedback.  What else should it do?

Sell More By Not Selling at All

Sometimes the most effective kinds of communication are the simplest.  And when it comes to reminding a prospect to take action, sometimes “no sell” is the best sell.  All it takes is a gentle reminder.   Here is an example from my other blog, Good Ad or Not.

The most effective reminders are ones that demonstrate an understanding of a customer or things that show you care.  If you’re in sales, find out what your prospects are up to and what they are working on.  And if they are working toward a hard deadline and you can help in any way, use those dates to check in and see if you can be a resource.  If a prospect has a presentation to give, learn more about the topic on your own time and send them some helpful content.  If they are stuck on a problem, offer to introduce them to an expert if you know one.  These helpful things brand you as a helpful person.  And they are subtle and gentle reminders of your company that will keep you top of mind.  So when it comes time for them to buy, you’ll be in the consideration set.  Because half of what you’re selling in sales is you.

The other day I interviewed a candidate for an open position.  She emailed me a thank you note and got an auto generated out of office message (which was still on from a recent trip), although I was in the office and she knew it.  She then emailed me a follow-up note saying “I noticed you still have your out of office message on.  If you don’t know how to turn it off here’s a link to instructions in Outlook in case should you need it.”  A gentle reminder was all it took.  I will be having her back for a second interview.

Sexy Uses of Frumpy Marketing Data

[tweetmeme source=bradmehl only_single=false]

Social media analytics is all the rage lately and for good reason. 

Compared to social media, which is very sexy now, email is old hat.  It’s downright frumpy — anything but sexy.   Many pundits have even predicted the death of email, given the rise of social media.  But for database marketers email is a channel and source of data with increasingly broad applications across the marketing spectrum.

This week I had the pleasure of meeting the winners of the Hearst Analytics Challenge, run by Charlie Swift, one of my former colleagues from LexisNexis.   Similar to the Netflix prize for the best movie recommendation algorithm, Hearst reached out to the analytics community and challenged them with developing a model to predict that variables that drive email results.   Smart.  And it got me thinking of other uses of email data.

Let’s take retail buying data.  Datalogix and other companies are taking offline transactional data like products purchased at retail (from providers like Nielsen) and matching it to online registration data — of which email address is a key data point.  If you take the retail purchase data, match it with registration/email data and associate it all with cookies then – poof – you’ve got the ability to serve targeted  online ads to people based on their offline behavior.  (After doing some deals with ad networks).   To make the sales of this data really scale, data companies are developing models of pre-packaged segments like organic food buyers.   So the soy milk company can serve online ads to people who are known buyers of organic foods.

How else can email data be used with other data sets?

Let’s take customer service interactions.   If a customer – let’s call him Joe – complains to Campbell’s that his soup is too salty he can be tagged and put into a “salt free” segment in their prospect database.   Hopefully when Joe called into the call center the rep captured his mail address  (or Campbell’s appended it by matching the incoming phone number to an email address).   If so, then Campbell’s could then run a predictive model to determine the attributes of “salt free” buying prospects like Joe and find look alikes from outside lists.  Then they would have email addresses of non-customers who may be prospects for its salt free soups.  All of this is doable with the right strategy.  Old hat for some marketers, new stuff to others.

Back to the online advertising piece.  If the email addresses are matched to online registration data, then couldn’t Campbell’s run online ads to other prospects — potential customers who are statistically more apt to buy a salt free soup?   And they could use the email addresses of known customers to build a profile over time with online behavioral data.  All they have to do is set cookies when customers click on the links and then collect other online data over time — like topics of interest from online interaction on their web site.  Then they can send Joe an email with helpful content like “foods that help lower blood pressure.”   He clicks on the link, goes to Campbell’s site and then views a recipe for clam chowder (assuming they have one).   Campbell’s could mine that data, connect the dots and then email or mail Joe a coupon for their salt free clam chowder.  By doing this, Campbell’s turns Joe from a complainer to an evangelist.

And how about product development.  If Campbell’s does not make a salt free clam chowder, mining the data could yield an insight:  there is a market for such a product.  Now that’s kind of sexy.  There is huge potential to use this kind of data in product development.

Email addresses change and surely email is not the only Rosetta Stone that can help join disparate data points, but in many cases it’s darn good. Add social media and CRM data in the mix and with proper analytics you’ve got got powerful insights that can elevate the whole customer experience, with the right applications and communication mix.

The key to email’s survival is relevance.  Marketers must be on target with emails that provide people with value above and beyond a transactional level.  If we do that then, well, email might just have more life left in it.

A Different Kind of Social Commerce

Social commerce is not new.  The term usually refers to sites like ShopSocially which enables retailers to turn their customers into ambassadors by asking friends shopping questions and sharing their purchase information. Or Blippy where users share updates about their online purchases with Facebook friends (and can integrate with sites like eBay, iTunes and Groupon).  But today I came across a different kind of social commerce.  Actually it was not social at all, it was a post in Amazon’s forum that was a crass blend of commerce with conversation.  Did it work (as a marketing tactic that would translate into sharing)?  This blog doesn’t have all the voting features of my other one, so I’ll simply link to it here. Let me know what you think.