Shadowy Data Matching Revealed in Graphic

Everyone who has followed me for any length of time knows that I generally despise “infographics.” They’re often sloppy about data. They tend to be visually cluttered and difficult to read. And they’re typically cynically created for SEO purposes.

However I like the Hubspot infographic below, even though it’s quite dense. It illustrates the shadowy world of data matching and trolling that goes on behind the scenes. Consumers have some vague high-level understanding this is going on but are ignorant of its extent, depth or precision.


My view is that several things need to happen in the market. To date “self-regulation” hasn’t worked. There probably now needs to be regulatory or legislative pressure to generate exposure and disclose of third party data sharing practices. Consumers must also be given some way to more directly participate, control or opt-out of this system (current “ad choices” is inadequate). In addition offline data collection and brokering must be brought into the discussion and made more visible as well.

Transparency and some measure of consumer control is the conceptual solution to the cat and mouse game of tracking and data brokering. Even though this Congress is incapable of passing legislation some form of regulation or legal action is coming around these issues. It will probably happen at the state level and/or via the FTC, creating uneven regulations and difficulty for all.

It would be better for the IAB, 4As and industry groups to come together and advocate true transparency.

Consumers will participate in data-sharing schemes if they have an understanding of how the data are used and they receive some benefit in return (personalization, convenience, offers, etc). However some of these practices would absolutely be shunned and condemned if consumers more knew about them.


Cosmic, Inexplicable Cross-Platform Spouse Retargeting on Facebook

Facebook ad targeting

This week I wrote about the Zillow acquisition of Trulia. I haven’t spent much time on either site for quite awhile. So last night I decided to go back and more thoroughly explore Trulia for the first time in a couple of years.

I performed a number of housing searches on my PC and looked at the features on the site. Subsequently, as you might imagine, I saw ads for Trulia on other sites on the PC. But what’s kind of amazing (or spooky) is what happened to my wife within literally minutes of me being on Trulia.

My wife was sitting across the room, absorbed in her Android Facebook app. (She’s quite an avid Facebook user — much more so than I am.) She knew I was on Trulia and so she told me that she was seeing ads for Trulia in her feed. What?

Either this was a cosmic coincidence — she wasn’t on Trulia at any point herself — or this was some kind of inexplicable cross-platform spouse retargeting.

Facebook knows that we’re married. I assume that she, by virtue of being my identified spouse, was receiving the same kinds of ads I would have seen on Facebook. I’ve never heard or seen anything like this. I’ve never heard of a retargeting capability that involves spouses or other closely associated persons: lovers, immediate relatives, etc.

Yet if there wasn’t some sort of spouse retargeting, how/why was she seeing these ads? What do you think this is? Where’s Dennis Yu?

Update: Somebody on Twitter just suggested to me that this was probably IP-retargeting. Both my wife and I were on our WiFi network. I’m sure that’s the explanation. Mystery solved.


Yelp Q2: 40% of Reviews Are from Mobile

Yelp logoYelp reported Q2 earnings this afternoon. It had a very strong quarter, beat Wall Street estimates and eked out its first first profit.

The company said that second quarter revenue was nearly $89 million, representing 61% growth over 2013. Net income was $2.7 million — the company’s first profit.

Total reviews now stand at 61 million. Roughly 40% of reviews are now generated from mobile devices according to the company. (It just added the capacity to post video.)

Total monthly uniques were 138 million with mobile 68 million mobile users. Yelp also reported that it now has just under 80K “active local business accounts” (revenue-generating accounts). The company operates in 27 countries.

For the full year Yelp projected revenues of between $372 million to $375 million.

Yelp Q2 2014

Yelp business accounts

Yelp Q2


Microsoft Is Now Likely Foursquare Buyer

Cortana + FoursquareIn my view it’s only a matter of time before Foursquare is acquired. We’re waiting for the company to launch its redesigned app but I’m skeptical that it will generate significant new momentum. I could be wrong.

I’ve always thought that Yahoo was the likely Foursquare buyer for several reasons. (Google might try to buy the company to keep it away from others but CEO Dennis Crowley probably would fight against a sale to Google.) However a new development suggests to me that Microsoft could well be the ultimate buyer of Foursquare. Indeed, those talks may already be happening.

Microsoft’s virtual assistant Cortana has started to incorporate Foursquare recommendations. Foursquare will power nearby personalized places recommendations.

Microsoft probably doesn’t have the immediate ability to replace this capability should Foursquare be acquired by Yahoo or Google. Although it might be able to strike a deal with Yahoo if it were the ultimate buyer. But the more that Foursquare is integrated and featured by Cortana the more important the property and its data become.

Thus I think that Microsoft now emerges as a very likely buyer of the company. Earlier it invested $15 million in Foursquare.

It would probably cost between $800 million and $1 billion to acquire Foursquare.

What do you think? Which of the following do you believe or tend to agree with:

  • The new app will return Foursquare to growth and lead to an IPO eventually
  • Foursquare will be bought by Yahoo, Google or Microsoft
  • It will be bought by someone else, perhaps a wild card like eBay or Alibaba

Image credit: Neowin, Reddit

Update: In the comments below, Andrew Shotland thoughtfully injects the name of Samsung into the speculative mix. I agree they’re a viable “other” in this conversation.

They want to build their own differentiated user experience. Now with the effective collapse of Tizen they may look to services like Foursquare to give them an edge as competition in the Android market continues to erode their dominance.

Update 2: Another suggestion is Apple. However my instinct is that Apple probably won’t buy Foursquare. But this list of 5 or so companies that may be interested could mean that Foursquare’s ultimate purchase price would probably be north of $1 billion and perhaps closer to $2 billion.


Coupon Searching on Mobile a Proxy for Immediate Purchase Intent

A new study from G/O Digital (a Gannett company) reveals some familiar and some interesting new things about mobile user behavior. The study, conducted in May, is noteworthy for its large sample: “over 13,000 responses from Key Ring app users . . . who own at least one desktop/laptop computer and a smartphone or tablet.”

Key Ring is a loyalty card app owned by Gannett.

The study sought to compare device usage in a broad range of shopping activities. The top “shopping” behavior performed on mobile devices was looking for coupons, followed by accessing shopping lists.

Key Ring top shopping activities mobile

While the chart above is not only about in-store mobile usage, prior surveys have documented the following top in-store mobile behaviors:

  1. Compare/checked prices
  2. Look for coupons or offers
  3. Search for product ratings/reviews

Depending on the survey the positions of the first two behaviors (coupons, price checking) may be reversed. The “access shopping list” behavior is quite interesting but not usually among the top 3 activities of in-store mobile users. Yet it’s an important piece of functionality for retail and especially grocery apps. It also ties in to indoor/in-store location, as I indicate below.

Below is a list of different behaviors (mostly cooking and grocery shopping-related), which reflect clear user device preferences:

Shopping activities by device

The survey also found certain kinds of consumer activities that can trigger purchases (“when and item goes on sale”) or are performed immediately prior to purchase (“search for offers”). As the following graphics reflect, sales motivate purchase behavior and searching for offers or coupons is often a proxy for immediate purchase intent:

purchase behavior Gannett

All this indicates how powerful coupons and sales are to motivate consumer purchase behavior. But coupons aren’t the only type of content that can motivate purchase behavior.

Agency BPN on behalf of Hillshire Brands found in an iBeacon case study (presented first at Place 2014) that product information/content and coupons equally impacted in-store purchases. In other words, brands don’t always need to use sales or deals to motivate consumers to buy.

Indeed, price is powerful but not always the determining factor in purchase decisions. Often reviews or item quality are the critical variables. In the apparel category, price outweighed all other factors by a significant margin. However in electronics or grocery price was a factor but other considerations also affected purchase behavior.

Key ring purchase variables

The report concludes with the assertion that shopping list-making “facilitates” in-store purchase behavior. Nearly 90% of survey respondents said that they were most inclined to make lists on smartphones vs. other devices. And 61% said that they do “often” or “always” prior to shopping.

The Gannett/Key Ring statement implies that lists by themselves impact in-store shopping. Lists are helpful but they don’t necessarily influence in-store behavior. They help organize and make it convenient. However when lists are augmented (in an app) by indoor location or related offers then they can become tools to influence in-store behavior.

Data on listmaking

Though this wasn’t directly addressed by the study, most retailer apps to date fail to deliver against mobile users’ basic in-store needs and behaviors. As I’ve argued many times retail and grocery apps need to become context-sensitive shopping assistants that can be used both away from stores (for e-commerce) or in-stores to facilitate purchases.

Once in stores retail apps should transform into sources of product information, product locations and offers (if available) to help consumers make purchase decisions, find products on the shelves or expose them to new products they might not have known about.

The metaphor for the ongoing disconnect between most retailers and consumers is the fact that 90%+ of coupons are still distributed in print.


Zillow’s Success Based on “Big Data”

ZillowZillow formally announced this morning that it was buying Trulia in an all-stock transaction worth $3.5 billion. I have followed both of these companies — less so after their IPOs — since their pre-launch days. (Makes me feel old.)

I recall my initial pre-launch meeting with Zillow’s then CEO Rich Barton. There were numerous real estate sites at the time. The key differentiator for Zillow was its massive aggregation of housing sales and pricing data. It took all that data, long before the tired phrase “big data” was in vogue, and turned it into a valuation algorithm.

That was the site’s major proposition at the time of launch.

Zillow as able to provide a “Zestimate” of the value of a large percentage of homes in the US. This seemed like magic to people and sparked massive media coverage and curiosity across the US. There was all kinds of “real estate voyeurism” triggered as a result.

In addition to checking out the value of your house you couldn’t help investigating how much your neighbor’s house was worth. There was a mix of high and low at play: a dazzling display of the internet’s capabilities to aggregate and present information, combined with an appeal to somewhat baser human qualities (greed, envy). But it worked. This single idea — how much is my house worth; how much is my neighbor’s house worth — drove initial interest and massive traffic.

From that early foothold, Zillow was able to add features, more content and so on. Zillow’s Barton saw the site ultimately as a “media property.” The model diversified over time into services. However Zillow grew and grew until it became the dominant US real estate destination. Competitor Trulia did almost as well.

Though less splashy than Zillow, Trulia was more of a utilitarian search engine for real estate listings with all kinds of related content. It executed extremely well against that idea and built the number two real estate property in the US. Now the two will be part of the same company, though Trulia will retain its brand and current leadership after the deal closes.

While Zillow’s leadership did a great job improving and evolving the site as real-estate market conditions changed, it was that initial big data “hook” of home valuations that propelled the site. Trulia offered a better user experience than most other sites and continued to build on that.

Both took two slightly different paths to their IPOs and now, with this merger, they’ll be part of a single real-estate holding company. But it was “big data” before big data became fashionable that is largely responsible for Zillow’s success.


Local SERP in ‘Turmoil’ after Google ‘Pigeon’ Algorithm Update

google-maps-iphone-iconSearch Engine Land reported last night that Google has implemented a new algorithm that in some cases radically changes the look and feel of the SERP for local queries. SEL has named the update “pigeon” (Google hasn’t disclosed a name).

Here’s what Google told SEL:

Google told us that the new local search algorithm ties deeper into their web search capabilities, including the hundreds of ranking signals they use in web search along with search features such as Knowledge Graph, spelling correction, synonyms and more.

In addition, Google said that this new algorithm improves their distance and location ranking parameters.

The new algorithm is currently rolling out for US English results and aims to provide a more useful and relevant experience for searchers seeking local results. Google didn’t share any details about if and when the update would roll out more widely in other countries and languages.

Mike Blumenthal this morning noticed missing 7 packs for various searches that otherwise would have shown them. DUI lawyers and local realtors were two that would have triggered 7 packs in the past.

Today there are many such searches without 7 packs it appears. However if there’s a geo-modifier in the query, the 7 pack did show up (according to various searches I performed). As an example, compare “DUI lawyers” and “SF DUI lawyer” below:

Pigeon SERP

Pigeon update

However the query “chiropractor,” without a geo-modifier, brought up the 7 pack for me. Perhaps it was based on the fact that it was one of a series of searches I had conducted for San Francisco.

Pigeon update

Beyond this many searches continue to bring up the carousel. For example see, “sushi, New York” or “movies, Los Angeles.”


As Matt McGee points out directory sites (e.g., Yelp) appear to have been boosted somewhat by this update. Many were demoted in the past in favor of individual business listings (7 pack).

The SERPs that are showing up are quite varied, depending on the business category and presence or absence of a modifier. Local SEOs appear to be having both positive and negative reactions — but overall confusion and turmoil have momentarily set in.

Here are some relevant links and discussion:

Update: some people are saying that the 7-pack is now MORE prominent than it was. In addition someone noticed businesses without physical addresses appearing in the 7-pack.

Please let me know if you see any particularly good analyses or discussions of the Pigeon local-algo update.

Update 2: This via Mike Blumenthal is from Moz. It shows the “local pack” (7-pack) dropping relative to other types of SERPs. It’s reportedly based on a review of thousands of localized queries. 

Moz Pigeon


The New Foursquare: Rebirth or Last Gasp?

New 4SQThe new Foursquare app is coming. And, depending on the reaction, it represents new life for the company or a last-chance prelude to a sale. My guess is probably the latter but I hope for the company’s sake it’s the former.

Yelp is 10 years old and Foursquare is half that old, having launched at SXSW as a sort of location-based game (collect badges, become the mayor). Today the company characterizes itself as a “50,000,000-strong community.”

The question that has never been satisfactorily answered, however, is how many active users are there? Anecdotal evidence suggests that usage has plateaued.

In its email letter this week to registered Foursquare users the company encouraged people to download Swarm, the new app that now houses check-ins. It says that “three quarters of you” are already on the new app (this is probably active users). It also previewed the new Foursquare app (and logo: “mix of map pin and superhero emblem”).

The company also provided a glimpse of how it will try and stand out and differentiate from Yelp, Google, TripAdvisor and others — in a word: personalization. It will offer different recommendations for everyone based on history, tastes, context, connections and more:

Everyone explores the world differently – guided by their own unique tastes, their friends, and the people they trust. Local search has never been good at this. It doesn’t get you, and, as a result, everyone gets the same one-size-fits-all results.

Why should two very different people get the same recommendations when they visit Paris? Or the same list of places when they’re looking for a bar? We’re about to change that. In a couple weeks, we’re rolling out a brand new version of Foursquare that’s all about you. Tell us what you like, and we’ll be on the lookout for great places that match your tastes, wherever you are . . .

No two people view the world exactly the same, so no two people will have the same experience with the app. 

The home screen will be the place that Foursquare’s new “personalized local search” (and local discovery) most obviously expresses itself. But sophisticated algorithms will also rank category results based on numerous signals that include individual history (“all of your check-ins and history will continue to help shape your recommendations”).

This personalization strategy is both smart and about the only major card Foursquare had to play. Execution will not be easy and harder still will be creating an overall UI/UX that is sufficiently familiar but simultaneously different enough so that users have affirmative reasons to use it rather than Yelp, OpenTable, TripAdvisor, YP or Google.

I still think the endgame for Foursquare is an acquisition. Would be buyers now include Microsoft, Yahoo, Apple and maybe even Google. I’m sure there are potential others too. In other words Foursquare will find a buyer; the question is how much.

I know that CEO Dennis Crowley is very ambitious and wants the company to realize its potential. However investors are not as idealistic or patient and will likely start agitating for a sale if there isn’t substantial traction within six months of launch.


God-Zillia: Trulia-Zillow Combo Would Dominate Online Real Estate

Trulia logoZillow is reportedly in talks to buy its biggest rival Trulia for roughly $2 billion according to Bloomberg. Both are public, though Zillow is the more valuable company by roughly 3X.

Zillow and Trulia are the number one and two US real estate sites respectively. Depending on whose list you consult either Yahoo Homes or are next. is operated by the National Assn of Realtors and would be the only remaining true competition to a “Zillia” combination. (Yahoo Homes is a distributor of Zillow content and advertising.)

Trulia value

According to June comScore data, cited by Bloomberg, together Zillow and Trulia represented “about 89 percent of all traffic to the 15 most-visited real estate sites.” Bloomberg also says the deal could be announced next week.

Zillow would likely be the surviving brand unless the company decides to maintain the Trulia site as a subsidiary entity. There’s an argument for doing that although it’s less efficient for the company to do so. 

Assuming the rumors are true, it will now probably cost Zillow more than $2 billion for Trulia. Both companies’ shares have risen on the takeover rumors.


Southwest Airlines Removes Passenger for Tweet about Bad Customer Service

Twitter logoYou may have read the story about the Minnesota man and his kids who were removed from a Southwest Airlines flight because he expressed dissatisfaction with the company’s boarding policy on Twitter.

The father wanted his two kids to board early with him but because of their status that was denied. Here’s USAToday’s summary of the incident:

Duff Watson of New Hope, Minn. [said] he’s an elite “A-List” member of Southwest’s Rapid Rewards frequent-flier program. A-List customers are entitled to early boarding because of their status, and Watson assumed that would include his children – ages 9 and 6 – who were traveling with him on a flight from Denver to Minneapolis.

A gate agent informed Watson that was not the case, and they’d have to wait until after priority boarding to get on the aircraft. But Watson, [reported he had] been able to board early with his kids on previous trips, claims the gate agent “rudely and dismissively” told him they’d have to wait.

The father, Duff Watson, expressed his frustration on Twitter and included the gate agent’s first name:


He was later asked to leave the flight because the gate agent felt “threatened.” Once he agreed to delete the tweet he and his children were allowed to get back on the flight.

The carrier, according to its own “contract of carriage” (embedded below) is permitted to deny boarding or remove passengers from the plane under a range of circumstances. Here’s my incomplete, edited version of the conditions or circumstances:

  • Whenever such action is necessary, with or without notice, for reasons of aviation safety
  • Force Majeure Event (outside carrier’s control)
  • Government Request or Regulation (compliance with law)
  • Interference with Flight Crew: Passengers who interfere or attempt to interfere with any member of the flight crew in carrying out its duties
  • Search of Passenger or Property
  • (Refusal to provide) Proof of Identity
  • Incompatible Medical Requirements
  • Comfort and Safety . . . including (i) Persons whose conduct is or has been known to be disorderly, abusive, offensive, threatening, intimidating, violent, or whose clothing is lewd, obscene, or patently offensive.
  • Weapons
  • Non-Smoking Policy (violation of)

I’ve highlighted the the clauses that Southwest might have been able to point to to justify removal of the passenger and his family for the remarks on Twitter. This is probably why the rude gate agent used the phrase “threatened.” Otherwise the guy’s conduct doesn’t qualify.

In addition Southwest’s specific boarding-related policies (at least available on its site) are largely silent on this situation: letting lower-status children board with a priority boarding family member (United allows this). The one relevant statement I could find was the following:

Can groups assigned to different boarding positions board together?

Yes. However, in order to maintain the integrity of the boarding process, we ask that earlier boarding positions board with the later positions. For example, if a passenger is assigned position A16 and wants to board with a passenger assigned position A45, the passenger holding the A16 boarding pass should board with the A45 passenger.

Watson’s Tweet clearly was not a threat to the gate agent, nor did it create an unsafe or disorderly condition on the plane. It was a customer service complaint. Accordingly Southwest breached its contract with Watson and his children. It also violated his First Amendment rights.

The airline deserves to be sued. However because he was ultimately let back on the flight there are probably no damages and the case isn’t worth filing.

The larger point, and why I find this so outrageous, is that there’s a power struggle going on between customers/consumers and corporations. Companies have for years gotten away with unfair policies and poor customer service and social media are now leveling the playing field and putting pressure on them. Some companies are reacting with nefarious terms or by using intimidation tactics. This incident was an example of the latter.

I don’t believe that agent really felt threatened. She probably felt angry and offended.

Many (not all) airline employees have a tense, condescending or openly antagonistic relationship with passengers. This was likely an act of hostility or retaliation by the Southwest employee.

Southwest needs to discipline and train the employee. It also needs to be much more communicative about its boarding policies. The right thing to do in this situation would have been to let the kids board with their father and inform him of the policy and that this was an exception. Instead the company (and its embittered employee) chose to be punitive.