Why, in an industry in which customer service is one of only two ways to set yourself apart, have brokers been so wary of measuring customer satisfaction?
It isn’t because agents don’t care. I’ve never met a Realtor who didn’t care about his customers. In fact, when we talked about Redfin’s customer satisfaction goals last week, a broker immediately contacted us to ask how we measure customer satisfaction.
Will the Brokers Who Measure Customer Satisfaction Please Step Forward?
I was going to refer him to someone else in real estate who uses our scoring system — we’re just beginners — but then I couldn’t find anyone else. Which is surprising, given that it’s the same system used by Apple, Costco, FedEx, American Express, Dell, Vanguard and literally thousands of other companies in virtually every industry — except real estate.
The broker ended up scheduling a meeting with Redfin’s Matt Goyer at Inman Connect. It may have been one of the only Connect meetings explicitly about customer satisfaction all week. In looking over the otherwise dazzling agenda — I have been to both Inman and Bloodhound conferences, and both are very good — I noticed that every type of marketing under the sun was in the program — except the one that works best: customer marketing.
The reason measuring customer satisfaction wasn’t more prominent is because people think of customer testimonials as the old-school word-of-mouth that Realtors traditionally rely on, whereas Connect is all about the new school: YouTube, Facebook and Twitter.
E-Commerce for Agents Not Houses
But the message matters as much as the new-fangled medium, and our message to the world has to be about the quality of our service. The same transformation in how consumers research listings will change how they choose an agent. Rather than meeting face to face to review listings, consumers now evaluate listings online. They can’t see or smell the house in person, so they bury their nose in numbers. When their first encounter with an agent is online, they’ll take the same approach.
Therein lies the great fallacy in many people’s original conception of real estate e-commerce: that the product being sold was a house. I’ve been asked a thousand times, “who would buy a house online?” In fact, what customers want to buy online isn’t a house but the services of a real estate agent.
Almost all of our customers don’t like choosing an agent based on intangibles. Who wants to be the judge in a personality contest? What people want is to choose an honest, competent, careful agent based on simple, objective criteria. We think the most important of these criteria will be customer satisfaction. The sites that have been scrambling to bring together all sorts of data about a property now need to do the same for an agent.
In this coming arms race, we believe that the most important weapon will be data about your own customers. And the only way you can use it is if you share all the data, not just a few cherry-picked reviews.
Gathering Customer Data
Since Redfin wrote up its first offer, we’ve surveyed everyone who works with us using a free online survey tool like Zoomerang. Zoomerang is dead-simple to use; no programming skill is required. We now have a formidable data, of thousands of survey responses, which we can use to guide not only our own decisions, but those of our clients.
The one question we’ve consistently asked in every survey is dead-simple too: whether the client would recommend the company to a friend. The yes/no answer to this question was from the beginning the primary determinant of our agents’ bonuses.
The Ultimate Question
Over the past year or so, after one of our board members read Frank Reichheld’s The Ultimate Question, we got a little more fancy about it. We asked clients the same question — it turns out that this question is Reichheld’s “ultimate” question — but with the answer rated on a scale of 1 – 10.
That allowed us to calculate what Reichheld calls a Net Promoter Score (NPS), which is the percentage of promoters (9’s and 10’s) minus the percentage of detractors (0 through 6). Everyone in customer service at Redfin, including me, has a bonus that depends on this score.
It turns out that the difference between just saying you’re “satisfied” and actually being a 10 out of 10 is huge. And we need as many 10’s as we can get: nobody is going to make the leap from seeing your web profile to becoming a client unless you have raving customer lunatics spreading your gospel all over the Web.
On the flip side, the damage one detractor can do to Redfin — I feel like I know them all by name — has been staggering. Given customers’ high expectations from real estate agents, it makes sense for us to add up the promoters and then dong ourselves for every detractor. Every one else in between affects our score too — as a lost opportunity to create a fanatic.
Raising Response Rates
Who gets surveyed? Anyone and everyone. We survey people who tour a property or make an offer, regardless of whether their offer goes through. About half respond. We used to give people a $25 Amazon gift certificate for filling out the survey, but we tried taking away the gift certificate and response rates didn’t change.
Reichhfeld’s book tells you to stop asking any questions in the survey except the ultimate question, so more people are likely to respond, but doing that never changed our response rate much. 50% – 60% of folks answer the question, while only 25% overall answer additional questions in a second survey, which is linked to the first.
The only tactic that really increased response rates was a short survey link, to prevent the link from breaking up across two lines in the email we use to send it out. We use bit.ly to shorten the link.
But Why Can’t the Clients Just Say It To Your Face?
By now, you’re probably thinking you want to do this informally, just in person whenever you get the chance. Big mistake.
Clients will tell you plenty in a survey that they would never have the courage to say to our face, especially when someone besides the agent — the broker, for example — sends out the survey, and anonymous responses are allowed. By rigorously surveying everyone, you can track down people you had a fleeting encounter with, and never find out why they chose another broker.
Letting that person just disappear is a lost opportunity — not just to earn a commission, but to get better. Our surveys have taught us not only how individual agents can raise their game — each agent gets a complete report on his client’s responses, every month — but how we have had to improve as a company: speeding up the website, hosting more tours and offering more agent choice.
We have made so many mistakes at Redfin that we can no longer hope to be smarter than our competition, but we do think we know more about our customers. I often wonder how a large brokerage has any idea what’s going on between its agents and its clients.
What We Like About NPS
To start learning, you could probably use any old survey and it would do the trick. But here’s what we like about measuring Net Promoter Score:
1. It’s simple. The best thing we ever did to our compensation plan was to simplify it. Pay through the nose for promoters; nothing for detractors.
2. It allows us to benchmark our performance to those of other companies. Anything above 50% is considered a good NPS score. You can do better.
3. It separates the goats from the sheep: simply asking a customer if he is satisfied resulted in uniformly high scores. Focusing on 9s and 10s makes it easier to reward the top performers.
4. Agents love it: getting a score every month that isn’t just a sales number, but a real reflection of how your customers feel about you is really uplifting. Agents need more feedback like that.
5. Consumers love it: we started publishing ratings to our website about halfway through 2008. Traffic to those pages shot up.
The only problem was that consumers don’t understand NPS well enough for us to publish our results.
This was easy enough to fix: in the survey we added the word “yes” below 6 and up, and “no” below 5 and on down. Now we can say that 97% of our clients would recommend our service — 97% gave us a rating that was correlated with “yes” — even though our NPS score for last quarter was only 75%. We can also publish the numeric rating for each transaction. Maybe as NPS becomes a more prevalent concept, we can publish our aggregate NPS score straight up. We’ll probably start doing that soon.
How Do You Measure Customer Satisfaction?
By now this post is very long, and we should cede the floor so we can hear how other brokers monitor customer satisfaction. All of us can learn from the discussion. In a time of economic anxiety, digital personas and skepticism about Realtors, letting the world know how we treat our customers is the only way to build trust, which is the only way to begin all the other conversations we want to have via Twitter, YouTube, Facebook and blogs.