Will adding a citizenship question to the Census harm the Market Research Industry?

The US Supreme Court appears likely to allow the Department of Commerce to reinstate a citizenship question on the 2020 Census. This is largely viewed as a political controversy at the moment. The inclusion of a citizenship question has proven to dampen response rates among non-citizens, who tend to be people of color. The result will be gains in representation for Republicans at the expense of Democrats (political district lines are redrawn every 10 years as a result of the Census). Federal funding will likely decrease for states with large immigrant populations.

It should be noted that the Census bureau itself has come out against this change, arguing that it will result in an undercount of about 6.5 million people. Yet, the administration has pressed forward and has not committed funds needed by the Census Bureau to fully research the implications. The concern isn’t just about non-response from non-citizens. In tests done by the Census Bureau, non-citizens are also more likely to inaccurately respond to this question than citizens, meaning the resulting data will be inaccurate.

Clearly this is a hot-button political issue. However, there is not much talk of how this change may affect research. Census data are used to calibrate most research studies in the US, including academic research, social surveys, and consumer market research. Changes to the Census may have profound effects on data quality.

The Census serves as a hidden backbone for most research studies whether researchers or clients realize it or not. Census information helps us make our data representative. In a business climate that is becoming more and more data-driven the implications of an inaccurate Census are potentially dire.

We should be primarily concerned that the Census is accurate regardless of the political implications. Adding questions that temper response will not help accuracy. Errors in the Census have a tendency to become magnified in research. For example, in new product research it is common to project study data from about a thousand respondents to a universe of millions of potential consumers. Even a small error in the Census numbers can lead businesses to make erroneous investments. These errors create inefficiencies that reverberate throughout the economy. Political concerns aside, US businesses undoubtably suffer from a flawed Census. Marketing becomes less efficient.

All is not lost though. We can make a strong case that there are better, less costly ways to conduct the Census. Methodologists have long suggested that a sampling approach would be more accurate than the current attempt at enumeration. This may never happen for the decennial Census because the Census methodology is encoded in the US Constitution and it might take an amendment to change it.

So, what will happen if this change is made? I suspect that market research firms will switch to using data that come from the Census’ survey programs, such as the American Community Survey (ACS). Researchers will rely less on the actual decennial census. In fact, many research firms already use the ACS rather than the decennial census (and the ACS currently contains the citizenship question).

The Census bureau will find ways to correct for resulting error, and to be honest, this may not be too difficult from a methodological standpoint. Business will adjust because there will be economic benefits to learning how to deal with a flawed Census, but in the end, this change will take some time for the research industry to address. Figuring things like this out is what good researchers do. While it is unfortunate that this change looks likely to be made, its implications are likely more consequential politically than it will be to the research field.

Long Live the Focus Group!

Market research has changed over the past two decades. Telephone research has faded away, mail studies are rarely considered, and younger researchers have likely never conducted a central location test in a mall. However, there is an old-school type of research that has largely survived this upheaval:  the traditional, in-person focus group.

There has been extensive technological progress in qualitative research. We can now conduct groups entirely online, in real-time, with participants around the globe. We can conduct bulletin board style online groups that take place over days. Respondents can respond via text or live video, can upload assignments we give them, and can take part in their own homes or workplaces. We can intercept them when they enter a store and gather insights “in the moment.” We even use technology to help make sense of the results, as text analytics has come a long way and is starting to prove its use in market research.

These new, online qualitative approaches are very useful. They save on travel costs, can be done quickly, and are often less expensive than traditional focus groups. But we have found that they are not a substitute for traditional focus groups, at least not in the way that online surveys have substituted for telephone surveys. Instead, online qualitative techniques are new tools that can do new things, but traditional focus groups are still the preferred method for many projects.

There is just no real substitute for the traditional focus group that allows clients to see actual customers interact around their product or issue. In some ways, as our world has become more digital traditional focus groups provide a rare opportunity to see and hear from customers. They are often the closest clients get to actually seeing their customers in a live setting.

I’ve attended hundreds of focus groups. I used to think that the key to a successful focus group was the skill of the moderator followed by a cleverly designed question guide. Clients spend a lot of time on the question guide. But they spend very little time on something that is critical to every group’s success: the proper screening of participants.

Seating the right participants is every bit as important as constructing a good question guide. Yet, screening is given passing attention by researchers and clients. Typically, once we decide to conduct groups a screener is turned around within a day because we need to get moving on the recruitment. In contrast, a discussion guide is usually developed over a full week or two.

Developing an outstanding screener starts by having a clear sense of objectives. What decisions are being made as a result of the project? Who is making them? What is already known? How will the decision path differ based on what we find? I am always surprised that in probably half of our qualitative projects our clients don’t have answers to these questions.

Next, it is important to remind clients that focus groups are qualitative research and we shouldn’t be attempting to gather a “representative” sample. Focus groups happen with a limited number of participants in a handful of cities and we shouldn’t be trying to project findings to a larger audience. If that is needed, a follow-up quantitative phase is required. Instead, in groups we are trying to delve deeply into motivations, explore ideas, and develop with new hypotheses we can test later.

It is a common mistake to try to involve enough participants to make findings “valid.” This is important, as we are looking for thoughtful participants and not necessarily “typical” customers. We want folks that will expand our knowledge of a subject and of customers will help us explore deeply into topics and develop new lines of inquiry we haven’t considered.

“Representative” participants can be quiet and reserved and not necessarily useful to this phase of research. For this reason, we always use articulation screening questions which raise the odds that we will get a talkative participant who enjoys sharing his/her opinions.

An important part of the screening process is determining how to segment the groups. It is almost never a good idea to hold all of your sessions with the same audience. We tend to segment on age, potentially gender, and often by the participants’ experience level with the product or issue. Contrasting findings from these groups is often where the key qualitative insights lie.

It is also necessary to over-recruit. Most researchers overrecruit to protect against participants who fail to show up to the sessions. We do it for another reason. We like to have a couple of extra participants in the waiting area. Before the groups start, the moderator spends some time with them. This accomplishes two things. First, the groups are off and running the moment participants enter the focus group room because a rapport with the moderator has been established. Second, spending a few minutes with participants before groups begin allows the moderator to determine in advance which participants are going to be quiet or difficult, and allows us to pay them the incentive and send them home.

Clients tend to insist on group sizes that are too large. I have viewed groups with as many as 12 respondents. Even in a two-hour session, the average participant will be talking for just 10 minutes in this case and that is if there are no silences or the moderator doesn’t talk! In reality, with 12 participants you will get maybe five minutes out of each one. How is that useful?

Group dynamics are different in smaller groups. We like to target having about six participants. This group size is small enough that all must participate and engage, but large enough to get a diversity of views.  We also prefer to have groups run for 90 minutes or less.

We like to schedule some downtime in between groups. The moderator needs this to recharge (and eat!), but this also gives time for a short debrief and to adjust the discussion guide on the fly. I have observed groups where the moderator is literally doing back-to-back sessions for six hours and it isn’t productive. Similarly, it is ideal to have a rest day in between cities to regroup to provide an opportunity to develop new questions. (Although, this is rarely done in practice.)

Clients also need to learn to leave the moderator alone for at least 30 minutes before the first group begins. Moderating is stressful, even for moderators who have led thousands of groups. They need time to review the guide and converse with the participants. Too many times, clients are peppering the moderator with last second changes to the guide and in general are stressing the moderator right before the first session. These discussions need to be held before focus group day.

We’d also caution against conducting too many groups. I remember working on a proposal many years ago when our qualitative director was suggesting we conduct 24 focus groups. She was genuinely angry at me when I asked her “what are we going to learn in that 24th group that we didn’t learn in the first 23?”.

In all candor, in my experience you learn about 80% of what you will learn in the first evening of groups. It is useful to conduct another evening or two to confirm what you have heard. But it is uncommon for a new insight arises after the first few groups. It is a rare project that needs more than about two cities’ worth of groups.

It is also critical to have the right people from the clients attending the sessions. With the right people present discussions behind the mirror become insightful and can be the most important part of the project. Too often, clients send just one or two people from the research team and the internal decision makers stay home. I have attended groups where the client hasn’t shown up at all and it is just the research supplier who is there. If the session isn’t important enough to send decision makers to attend, it probably isn’t important enough to be doing in the first place.

I have mixed feelings about live streaming sessions. This can be really expensive and watching the groups at home is not the same as being behind the mirror with your colleagues. Live streaming is definitely better than not watching them at all. But I would say about half the time our clients pay for live streaming nobody actually logs in to watch them.

Focus groups are often a lead-in to a quantitative study. We typically enter into the groups with an outline of the quantitative questionnaire at the ready. We listen purposefully at the sessions to determine how we need to refine our questionnaire. This is more effective than waiting for the qualitative to be over before starting the quantitative design. We can usually have the quant questionnaire ready for review before the report for the groups is available because we take this approach.

Finally, it is critical to debrief at the end of each evening. This is often skipped. Everyone is tired, has been sitting in the dark for hours, and have to get back to a hotel and get up early for a flight. But, a quick discussion to agree on the key takeaways while they are fresh in mind is very helpful. We try to get clients to agree to these debriefings before the groups are held.

Traditional groups provide more amazing moments and unexpected insights than any other research method. I think this may be why, despite all the new options for qualitative, clients are conducting just as many focus groups as ever.

The Importance of Employee Surveys

America’s CEOs constantly say “Our people are our most important asset.” But how many organizations actually live up to this credo?

When I hear a leader say this I’ll investigate where the head of HR reports. I often discover that the top HR person reports to the COO or the CFO. That is a warning sign, as a leader who truly believes in their workforce as a strategic asset would have HR reporting directly to the CEO.

Another warning sign that this credo may ring hollow is if the organization fails to conduct an annual employee survey. I’ve worked with organizations that spend millions of dollars each year on market research yet never get around to an employee survey. In many cases it is because they are worried about what they might discover and recognize that conducting an employee survey sets up an expectation that actions will result from it.

I have been involved in about 250 employee surveys. Results can be eye opening and help create an open culture in an organization. How they are conducted and their content can tell a lot about what the culture of the organization is like.

The best employee projects I have been involved with have the following characteristics:

  1. They have sincere support and commitment from the CEO and the head of HR. This is essential to an employee survey’s success.
  2. They are developed by researchers and not HR or HR consultants. HR-produced surveys tend to be based on amorphous, general concepts and results tend to not be actionable. Surveys written by research departments and firms tend to be more specific and provide more direction as to what to do with the results.
  3. They are conducted annually. Similarly, they are part of a long term, continuous process of listening to employees and making changes based on their feedback.
  4. Results are used in managerial evaluations but are given the proper weight in these performance reviews and are not over-emphasized.
  5. Results are openly shared with the staff. Not just some of the results, but everything that was asked.
  6. They are conducted with a third-party. An outside firm can be objective in analyzing the results and can place results in a context of other projects the have conducted.

But, by far the most important success criteria for these projects is employees must be confident that changes will happen as a result of the survey. We counsel clients to put our recommended actions into three categories:

  • Changes that can be made quickly and visibly. These are little, inexpensive things that should be done right away, and the staff should know that they were made as a result of the survey.
  • Important changes that will require time and effort. These are more substantive changes that may require more input from employees and investment to make happen. Employees should know these efforts are happening and ideally be a part of them.
  • Changes that are recommended but that leadership will choose not to make. This is a step often skipped. The survey will uncover changes that leadership isn’t prepared to make, that require too much investment, etc. It is important that leaders acknowledge these to the staff. It is often sufficient to mention that the survey uncovered these items, but at this time priority will not be given to them. Employees greatly appreciate this honesty.

By far the biggest mistake that can be made with an employee survey is to conduct it and then make no changes based on its results. In my experience, this happens at least 50% of the time. I have counseled dozens of potential clients away from conducting an employee study because I didn’t feel they were prepared to act on the results. There is nothing worse than asking your entire employee base for feedback, and then ignoring the feedback they provide.

Employee surveys can be an asset to any organization. I honestly wouldn’t recommend working anywhere that doesn’t conduct an annual survey and doesn’t make changes based on the results.

Is getting a driver’s license still a rite of passage for teens?

In the 80’s and 90’s, before the Millennial generation hit their teen years in force, we would use “driver’s license status” as a key classification variable in studies. Rather than split focus groups by age or grade in school, we would often place teens who had their license in one group and those who did not have their license yet in another group. Regardless of the topic of the group. We found that teens with licenses were more independent of their parents and more capable of making decisions without parental input. Drivers license obtention was often better predictor of consumer behavior than age.

Young people experience many rites of passages in a short period of time. These are experiences that signify a change in their development. They ride the school bus for the first time, get their first smartphone, enter high school, go to the prom, leave home to go to college, vote for the first time, etc. As marketers, we have always looked at these inflection points as times when consumer behavior shifts. The obtaining of a driver’s license is traditionally seen as a watershed moment as it signifies a new level of independence.

However, this wisdom no longer holds. Millennials, particularly second wave Millennials, are not as focused on obtaining drivers licenses as their Boomer and Xer parents were. Where I grew up, we couldn’t wait until our 16th birthday so we could get our learner’s permit. My classmates and I usually took our road tests at the first opportunity. Failing the road test was a traumatic experience, as it caused us to remain in our parents’ control for a few more months.

This is no longer the case. In 1983, 46% of America’s 16-year-olds had a driver’s license. That is now less than 25% currently. I was very surprised to notice that my children and their friends seemed to be in no particular rush to get their licenses. Many times, it was the parents that pushed the kids to take their road test, as the parents were tiring of chaperoning the kids from place to place.

There are likely things that have caused this change:

  • Today’s parents are highly protective of children. Parents no longer push their children to be as independent as quickly.
  • There are societal pressures. In most states, there are more stringent requirements in terms of driving experience to be able to take a road test and more restrictions on what a younger driver can do with his/her license. The license simply isn’t as valuable as it used to be.
  • Driving has peaked in the US. People are driving less frequently and fewer miles when they do. There has also been a movement of the population to urban areas which have more mass transit.
  • The decline of retail has played a part. Going to the mall was a common weekend activity for Xer teens. Now, staying home and shopping on Amazon is more common. Millennials never went to the mall to socialize.
  • Online entertainment options have proliferated. Movies and shows are readily streamed. Many teens fulfill a need for socialization via gaming, where they interact with their friends and make new ones. This need could only be met in person in the past.
  • Teens are working less so have less of a need to drive to work. Of course, this means they have less of their own money and that tethers them to their parents even longer.

There are likely many other causes. But the result is clear. Teens are getting licenses later and using them less than they did a generation ago.

As a result, researchers have lost a perfectly good measure! Obtaining a driver’s license is not as strong a rite of passage as it used to be.

We’ve been thinking about what might make a good alternative measure. What life event do young people experience that changes them in terms of granting their independence from parents? Leaving home and living independently for the first time would qualify but seems a bit late to be useful. There may be no clear marker signifying independence for Millennials, as they stay dependent on parents across a much wider time period than in the past. Or, perhaps we need to change our definition of independence.

Why Lori is the Best Shark in the Tank

Shark Tank is one of my favorite TV shows. It showcases aspiring entrepreneurs as they make business presentations to an investor panel, who then choose whether to invest. It is fun to play along and try to predict how the Sharks will react to a business pitch. As a small business owner/entrepreneur, it is fun to imagine how I might do in a Shark Tank presentation. And, it was interesting to watch a college teammate of mine make a successful pitch on the show.

My inner geek came out when watching a recent episode. I got to wondering how much the need to entertain might cloud how the venture capital world is portrayed on the show. How many Shark Tank pitches actually result in successful companies? Is the success rate for Shark Tank businesses any higher than any other small company looking for growth capital? Are there any biases in the way the Shark Tanks choose to invest?

This curiosity led to a wasted work day.

Venture capital, especially at early stages, involves high risk bets. Firms may invest in 100 companies knowing full well that 80 or 90 of them will fail, but that a handful of wild successes will pay off handsomely. It isn’t for the faint of heart. I found an interview with Mark Cuban where he stated he hoped that 15% of his Shark Tank investments would eventually pay off. Even that seems high. Given that he has invested about $32 million so far, that is an admission that $27.5 million of that is expected to be wasted. Gutsy.

I also was able to discover interesting things about the show that are largely hidden from the viewer:

  1. The Sharks themselves are paid to take part. I was able to find discussions that suggested they may make as much as $100K per episode. That is a million dollars or more per season, so perhaps they are playing with house money more than they let on.
  2. Getting on Shark Tank is statistically harder than getting into an Ivy League college. It is estimated that more than 50,000 people apply for each season with less than 1% being successful. That alone should provide some realism as to the probability of success of new businesses.
  3. In the early seasons, an entrepreneur had to give up 2% of revenue or 5% of his/her company to the production firm just to appear on the show. That requirement was removed in later seasons because Mark Cuban refused to remain on the show if it remained.
  4. Many of the deals you see made on the show don’t end up being consummated. Forbes conducted survey research in 2016 that indicated that 43% of Shark Tank deals fell apart in the due diligence stage and 30% of the time the deal changed substantially from what is seen on TV. The deal you see on TV only came to fruition as you saw it about 1 in 4 times (27% of the time).

This makes it challenging to assess the deals and whether or not they paid off. Shark Tank companies are almost all privately-held so their revenue data is tough to come by and we can’t really know for sure what the deal was.

Although we can’t review business outcomes as we might like, we can look closely at the deals themselves. The data we used for this includes all deals and prospective deals from the first nine seasons of the show. So, it does not include the current season, which premiered in October 2018.

In the first nine seasons, there were 803 pitches resulting in 436 closed on-air deals (53% of pitches). Applying the Forbes data would imply that of these 436 deals, 187 of them likely feel apart, and 131 of them likely changed substantially. The net? Our projection would be that 53% of pitches result in handshakes on-air, but post-show only 37% of all original pitches close at all and only 15% of pitches will close at the terms you see on air.

Why would Shark Tank deals fail to close? There is a due diligence stage where investors get to have their accountants review the entrepreneur’s books. I found some articles that indicated that some entrepreneurs got cold feet and refused the deal after the show. Also, some of the deals have contingencies which fail to occur.

It is interesting to look at deals by the gender of the entrepreneur as it shows that Shark Tank entrepreneurs skew heavily male:

  • Men are much more likely than women to appear as entrepreneurs on Shark Tank. Of the 803 pitches, 482 (60%) made by men, 198 (25%) by women, and 119 (15%) by mixed teams of men and women. So, 75% of the time, at least one male is involved in the pitch, and 40% of the time at least one female is involved in the pitch.
  • However, women (59% closed) are more likely than men (51%) to successfully close a deal on air.

There are data that imply that men and women negotiate differently:

  • Men initially ask for higher company valuations ($4.5 million on average) than women ($3.1 million on average).
  • Men also ask for more capital ($342K on average) than women ($238K on average).
  • Men (47%) and women (49%) receive about the same proportion of their initial valuation ask. Men (94%) and women (88%) also receive about the same proportion of cash that they initially ask for.

So, men are far more likely to appear on the show and come with bigger deals on average than women. But they receive (proportionately) about the same discount on their deals as women as they negotiate with the Sharks. If there is a difference in their negotiation skills it is that men start bigger or come to the show with more mature companies.

We can also look at individual Sharks to get a sense of how good of negotiators they are:

  • Mark is the most aggressive Shark. He has the most deals (132, despite not being on the early seasons of the show) as well as the most invested (about $32 million).
  • The cheapest (or most frugal?) Sharks are Barbara and Daymond. Barbara has put forth the least amount of money (about $10 million) and her average deal valuation is $945K. Daymond has put out the second least amount of money (about $12 million) and has an average deal size of $957K. These two Sharks have likely not put much more money into their Shark investments than they have been paid to be on the show.
  • Mr. Wonderful seems to have a “go big or stay home” mentality. He has closed the fewest deals (64) of any Shark. But, his average deal valuation of $2.7 million is the highest of any Shark.
  • Lori and Kevin (31% of pitches) are the most likely to make an offer. Barbara and Daymond (22%) are least likely to make an offer.
  • So, Kevin make the most offers and closes the fewest deals, making him the least desirable Shark from the standpoint of the entrepreneurs.

Barbara is the most likely to invest in a female entrepreneur. She is about as likely to invest in a female entrepreneur as a male entrepreneur despite the fact that so many more men than women appear on the show. Kevin and Robert are the least likely to invest in a female entrepreneur. Mark and Daymond demonstrate no bias, as the invest in about the same proportion as appearances on the show.

ALL

Barbara Lori Mark Daymond Kevin

Robert

Male

60%

44% 53% 60% 57% 67%

71%

Female

25%

42% 33% 27% 25% 19%

17%

Mixed Team

15%

14% 14% 13% 18% 14%

12%

So, who has been the most successful Shark? It can be hard to tell because data are scarce, but my vote would go for Lori. USA Today put out a list of the top 20 best-selling products featured on Shark Tank. Six of the top 10 were from investments Lori made, including the top 3. Eight of the top 10 investments by revenue were made by the two female Sharks, Lori and Barbara.

Who are the worst Sharks in terms of revenue performance? My vote here would be a tie between Mark and Daymond. Mark has just 3 of the top 20 investments and Daymond has just 2. If we can assume that the goal of venture capital is to generate big wins, it is clear that Lori and Barbara are killing it and Mark and Daymond are not.

Shark Tank is a great catalyst for entrepreneurs, but because it is entertainment and not reality it can mischaracterize entrepreneurship in the real world. Sharks may invest for the entertainment value of the show and because investing boosts their personal brand as much as the product. And, it might just be the case that the amount of money they have invested is not much larger than the amount of money they have been paid to be on the show.

Almost all successful people will tell you that learning from their failures was at least as important as their successes, yet Shark Tank never revisits failed investments and it is likely that the bulk of the deals we see on TV do not end up paying off for the investor. The show does not disclose how few of its deals actually come to fruition once the cameras are no longer rolling. Just once I’d like to see an update segment show an investment that failed miserably.

Shark Tank also seems to imply that hard work and grit always triumph, when in reality knowing when to cut losses and having a little bit of luck matters a lot in business success. Grit matters for sure, but not when it’s focus is blind and irrational, and it can be sad to see entrepreneurs who have sacrificed so much and it is clear their business is not going to make it.

At its best, Shark Tank stimulates people to think like an entrepreneur. At its worst, it presents too rosy a picture of small business life which influences people to invent new products and launch companies that are likely to fail, at great consequence to the entrepreneur. It certainly provides great entertainment.

How Did Pollsters Do in the Midterm Elections?

Our most read blog post was posted the morning after the 2016 Presidential election. It is a post we are proud of because it was composed in the haze of a shocking election result. While many were celebrating their side’s victory or in shock over their side’s losses, we mused about what the election result meant for the market research industry.

We predicted pollsters would become defensive and try to convince everyone that the polls really weren’t all that bad. In fact, the 2016 polls really weren’t. Predictions of the popular vote tended to be within a percent and a half or so of the actual result which was better than for the previous Presidential election in 2012. However, the concern we had about the 2016 polls wasn’t related to how close they were to the result. The issue we had was one of bias: 22 of the 25 final polls we found made an inaccurate prediction and almost every poll was off in the same direction. That is the very definition of bias in market research.

Suppose that you had 25 people flip a coin 100 times. On average, you’d expect 50% of the flips to be “heads.” But, if say, 48% of them were “heads” you shouldn’t be all that worried as that can happen. But, if 22 of the 25 people all had less than 50% heads you should worry that there was something wrong with the coins or they way they were flipped. That is, in essence, what happened in the 2016 election with the polls.

Anyway, this post is being composed the aftermath of the 2018 midterm elections. How did the pollsters do this time?

Let’s start with FiveThirtyEight.com. We like this site because they place probabilities around their predictions. Of course, this gives them plausible deniability when their prediction is incorrect, as probabilities are never 0% or 100%. (In 2016 they gave Donald Trump a 17% chance of winning and then defended their prediction.) But this organization looks at statistics in the right way.

Below is their final forecast and the actual result. Some results are still pending, but at this moment, this is how it shapes up.

  • Prediction: Republicans having 52 seats in the Senate. Result: It looks like Republicans will have 53 seats.
  • Prediction: Democrats holding 234 and Republicans holding 231 House seats. Result: It looks like Democrats will have 235 or 236 seats.
  • Prediction: Republicans holding 26 and Democrats holding 24 Governorships. Result: Republicans now hold 26 and Democrats hold 24 Governorships.

It looks like FiveThirtyEight.com nailed this one. We also reviewed a prediction market and state-level polls, and it seems that this time around the polls did a much better job in terms of making accurate predictions. (We must say that on election night, FiveThirtyEight’s predictions were all over the place when they were reporting in real time. But, as results settled, their pre-election forecast looked very good.)

So, why did polls seem to do so much better in 2018 than 2016? One reason is the errors cancel out when you look at large numbers of races. Sure, the polls predicted Democrats would have 234 seats, and that is roughly what they achieved. But, in how many of the 435 races did the polls make the right prediction? That is the relevant question, as it could be the case that the polls made a lot of bad predictions that compensated for each other in the total.

That is a challenging analysis to do because some races had a lot of polling, others did not, and some polls are more credible than others. A cursory look at the polls suggests that 2018 was a comeback victory for the pollsters. We did sense a bit of an over-prediction favoring the Republican Senatorial candidates, but on the House side there does not seem to be a clear bias.

So, what did the pollsters do differently? Not much really. Online sampling continues to evolve and get better, and the 2016 result has caused polling firms to concentrate more carefully on their sampling. One of the issues that may have caused the 2016 problem is that pollsters are starting to almost exclusively use the top 2 or 3 panel companies. Since 2016, there has been a consolidation among sample suppliers, and as a result, we are seeing less variance in polls as pollsters are largely all using the same sample sources. The same few companies provide virtually all the sample used by pollsters.

Another key difference was that turnout in the midterms was historically high. Polls are more accurate in high turnout races, as polls almost always survey many people who do not end up showing up on election day, particularly young people. However, there are large and growing demographic differences (age, gender, race/ethnicity) in supporters of each party, and that greatly complicates polling accuracy. Some demographic subgroups are far more likely than others to take part in a poll.

Pollsters are starting to get online polling right. A lot of the legacy firms in this space are still entrenched in the telephone polling world, have been protective of their aging methodologies, and have been slow to change. After nearly 20 years of online polling the upstarts have finally forced the bigger polling firms to question their approaches and to move to a world where telephone polling just doesn’t make a lot of sense. Also, many of the old guard, telephone polling experts are now retired or have passed on, and they have largely led the resistance to online polling.

Gerrymandering helps the pollster as well. It still remains the case that relatively few districts are competitive. Pew suggests that only 1 in 7 districts was competitive. You don’t have to be a pollster to accurately predict how about 85% of the races will turn out. Only about 65 of the 435 house races were truly at stake. If you just flipped a coin in those races, in total your prediction of house seats would have been fairly close.

Of course, pollsters may have just gotten lucky. We view that as unlikely though, as there were too many races. Unlike in 2018 though, in 2016 we haven’t seen any evidence of bias (in a statistical sense) in the direction of polling errors.

So, this is a good comeback success for the polling industry and should give us greater confidence for 2020. It is important that the research industry broadcasts this success. When pollsters have a bad day, like they did in 2016, it affects market research as well. Our clients lose confidence in our ability to provide accurate information. When the pollsters get it right, it helps the research industry as well.

The most selective colleges have the least effective marketing

Recently, Stanford University made headlines for deciding to stop issuing an annual press release documenting its number of applicants and acceptances.

There has been a bit of an arms race among colleges with competitive admissions to be able to claim just how selective they are. The smaller the proportion of applicants accepted, the better the college does in many ranking systems and the more exclusive the “brand” of the college becomes.

This seems to be a bit crazy, as publicizing how few students are accepted is basically broadcasting how inefficient your college marketing system has become. We can’t think of any organization beyond colleges that would even consider doing something analogous to this – broadcasting to the world that they have enticed non-qualified buyers to consider their product.

I learned firsthand how ingrained this behavior is among college admissions and marketing personnel. About five years ago I had the pleasure to speak in front of a group of about 200 college marketers and high school counselors. I created what I felt was a compelling and original talk which took on this issue. I have given perhaps 200 talks in my career, and this one might have been the single most poorly received presentation I have ever delivered.

The main thrust of my argument was that as a marketer, you want to be as targeted as possible so as to not waste resources. “Acquisition cost” is an important success metric for markers: how much do you spend in marketing for every customer you are able to obtain? Efficiency in obtaining customers is what effective marketing is all about.

I polled the audience to ask what they felt the ideal acceptance rate would be for their applicants. Almost all responded “under 10%” and most responded “under 5%.” I then stated that the ideal acceptance rate for applicants would be 100%. The ideal scenario would be this: every applicant to your college would be accepted, would then choose to attend your institution, would go on to graduate, become a success, and morph into an engaged alumnus.

I used an analogy of a car dealership. Incenting college marketers to increase applications is akin to compensating a car salesperson for how many test drives he/she takes customers on. The dealership derives no direct value from a test drive. Every test drive that does not result in a car purchase is a waste of resources. The test drive is a means to an end and car dealers don’t tend to track it as a success metric. Instead, they focus on what matters – how many cars are sold and how much was spent in marketing to make that happen.

Colleges reward their marketers to get students to test drive when they should be rewarding their marketers for getting them to buy. This wouldn’t matter much if a high proportion of applicants were accepted and ending up attending.  But, even at highly selective colleges it is not uncommon for less than 10% of applicants to be accepted, less than 33% of those accepted to choose to attend, and less than 50% of those that enroll to actually end up graduating. At those rates, for every 1,000 applicants, just 17 will end up graduating from the institution. That is a success rate of 1.7%.

These are metrics that in any business context would be seen as a sign of an organization in serious trouble. Can you imagine if only 10% of the people who came in your store qualified to buy your product? And then if only a third of those would actually decide to do so? And then if half of those that do buy don’t end up using your product or return it? That is pretty much what happens at selective colleges.

This issue is a failure of leadership. College marketers I have worked with can often see this problem, but feel pressured by their Deans and College Presidents to maximize their applicant base. Granted, this can help build the college’s brand, but it is a huge drain on resources that are better spent ensuring targeting applicants who are poised for success at the institution. It has happened because selectivity is considered important in building a college’s brand. Stanford has taken a useful first step, and hopefully other colleges will follow their lead.

Is segmentation just discrimination with an acceptable name?

A short time ago we posted a basic explanation of the Cambridge Analytica/Facebook scandal (which you can read here). In it, we stated that market segmentation and stereotyping are essentially the same thing. This presents an ethical quandary for marketers as almost every marketing organization makes heavy use of market segmentation.

To review, marketers place customers into segments so that they can better understand and serve them. Segmentation is at the essence of marketing. Segments can be created along any measurable dimension, but since almost all segments have a demographic component we will focus on that for this post.

It can be argued that segmentation and stereotyping are the same thing. Stereotyping is attaching perceived group characteristic to an individual. For instance, if you are older I might assume your political views lean conservative, since it is known that political views tend to be more conservative in older Americans that they are in general among younger Americans. If you are female I might assume you are more likely to be the primary shopper for your household, since females in total do more of the family shopping than males. If you are African-American, I might assume you have a higher likelihood than others to listen to rap music, since that genre indexes high among African-Americans.

These are all stereotypes. These examples can be shown to true of a larger group, but that doesn’t necessarily imply that they apply to all the individuals in the group. There are plenty of liberal older Americans, females who don’t shop at all, and African-Americans who can’t stand rap music.

Segmenting consumers (which is applying stereotypes) isn’t inherently a bad thing. It leads to customized products and better customer experiences. The potential problem isn’t with stereotyping, it is when doing so moves to a realm of being discriminatory that we have to be careful. As marketers we tread a fine line. Stereotyping oversimplifies the complexity of consumers by forming an easy to understand story. This is useful in some contexts and discriminatory in others.

Some examples are helpful. It can be shown that African-Americans have a lower life expectancy than Whites. A life insurance company could use this information to charge African-Americans higher premiums than Whites. (Indeed, many insurance companies used to do this until various court cases prevented them from doing so.) This is a segmentation practice that many would say crosses a line to become discriminatory.

In a similar vein, car insurance companies routinely charge higher risk groups (for example younger drivers and males) higher rates than others. That practice has held up as not being discriminatory from a legal standpoint, largely because the discrimination is not against a traditionally disaffected group.

At Crux, we work with college marketers to help them make better admissions offer decisions. Many colleges will document the characteristics of their admitted students who thrive and graduate in good standing. The goal is to profile these students and then look back at how they profiled as applicants. The resulting model can be used to make future admissions decisions. Prospective student segments are established that have high probabilities of success at the institution because they look like students known to be successful, and this knowledge is used to make informed admissions offer decisions.

However, this is a case where a segmentation can cross a line and become discriminatory. Suppose that the students who succeed at the institution tend to be rich, white, female, and from high performing high schools. By benchmarking future admissions offers against them, an algorithmic bias is created. Fewer minorities, males, and students from urban districts will be extended admissions offers What turns out to be a good model from a business standpoint ends up perpetuating a bias., and places certain demographics of students at a further disadvantage.

There is a burgeoning field in research known as “predictive analytics.” It allows data jockeys to use past data and artificial intelligence to make predictions on how consumers will react. It is currently mostly being used in media buying. Our view is it helps in media efficiency, but only if the future world can be counted on to behave like the past. Over-reliance on predictive analytics will result in marketers missing truly breakthrough trends. We don’t have to look further than the 2016 election to see how it can fail; many pollsters were basing their modeling on how voters had performed in the past and in the process missed a fundamental shift in voter behavior and made some very poor predictions.

That is perhaps an extreme case, but shows that segmentations can have unintended consequences. This can happen in consumer product marketing as well. Targeted advertising can become formulaic. Brands can decline distribution in certain outlets. Ultimately, the business can suffer and miss out on new trends.

Academics (most notably Kahneman and Tversky) have established that people naturally apply heuristics to decision making. These are “rules of thumb” that are often useful because they allow us to make decisions quickly. However, these academics have also demonstrated how the use of heuristics often result in sub-optimal and biased decision making.

This thinking applies to segmentation. Segmentation allows us to make marketing decisions quickly because we assume that individuals take on the characteristics of a larger group. But, it ignores the individual variability within the group, and often that is where the true marketing insight lies.

We see this all the time in the generational work we do. Yes, Millennials as a group tend to be a bit sheltered, yet confident and team-oriented. But this does not mean all of them fit the stereotype. In fact, odds are high that if you profile an individual from the Millennial generation, he/she will only exhibit a few of the characteristics commonly attributed to the generation. Taking the stereotype too literally can lead to poor decisions.

This is not to say that marketers shouldn’t segment their customers. This is a widespread practice that clearly leads to business results. But, they should do so considering the errors and biases applying segments can create, and think hard about whether this can unintentionally discriminate and, ultimately, harm the business in the long term.


Visit the Crux Research Website www.cruxresearch.com

Enter your email address to follow this blog and receive notifications of new posts by email.