Day 2 of AMEC’s Measurement Week NYC started off with Help a Reporter Out (HARO) founder Peter Shankman showing how measurement plays a role in everyday events and closed with Western Governors University (WGU) data scientist Matt Poulton discussing predictive analytics.
Let’s take a look at some highlights from those two presentations.
Nail Customer Loyalty: The Metrics That Matter
Content creation and data analysis are great, Peter Shankman says, but your core job is to generate revenue.
“Anything you do that does not have an end goal of generating revenue is wrong,” Peter says.
The key to generating revenue is all of the data that you have. If you, like many others, aren’t using it, you’re missing something. Here are some of Peter’s pieces of advice.
1. Brand everything
Back at the height of the film Titanic’s popularity, Peter sold 500 shirts that read “It sank. Get over it.” in six hours. He called USA Today, started selling them online and when the story broke, sold 10,000 more.
These days, he couldn’t have had that same success. Copycats would’ve taken his idea and cut into his sales. What can you do to limit that?
“Brand everything you do,” Peter says. “The only way to track data coming back to you is if it’s branded back to you.”
Peter learned the lesson the hard way when Lance Armstrong tweeted one of his videos. His Web traffic spiked, but the video didn’t have a brand call back. Peter admitted that he blew the opportunity to send people to his website, Twitter or other online properties.
2. Un-silo
If customer service gets 15 calls about a problem but has to fill out 15 forms, the issue will blow up when it could have easily been avoided.
Customer service should have an all access pass to engage anyone within your organization because they find out first what’s broken, Peter says.
3. Be transparent
No matter how careful you are, you will make a mistake. Be prepared to get in front of it.
“The only way to beat the haters back is with transparency,” Peter says.
When Barack Obama first took office, he faced backlash after Tom Daschle, his nomination for secretary of health and human services, was found to have tax errors. Obama’s response of “I screwed up” made it difficult for reporters to play hardball, Peter says.
When people try to run away from scandals or issues it can turn an eight-hour story into an eight-month story.
“At the end of the day, if you’re transparent, everything will work in your favor,” Peter says. “There’s no better lover in the world than a former hater.”
4. Be relevant
“We’ve never had more information than now, but we refuse to use it,” Peter says. “We have data on every single one of our customers even if it’s just an email address.”
This data can help us tailor experiences for our audiences.
When Peter operated HARO, he would send subscribers a monthly survey. He would then respond to each of the respondents personally and tell them if he was taking their suggestion.
When people saw that he took their suggestions, they became invested in the platform. Why does that matter? Invested customers spend 2.5 times more than others, Peter says.
5. Embrace concepts, not brand
Peter recommends that you design for the future but not a specific service. For example, brands that design with mobile in mind put themselves in a better spot than those who design with Twitter in mind.
Who knows what will happen to Twitter? It may not exist as we know it or at all in 2020. Concepts, though, take longer to change than brands.
6. Stay top-of-mind
“When you are top-of-mind, you give your audience all the reason in the world to give you data which you can measure,” Peter says
Former Paramount Pictures CEO Barry Diller got to the office early so he could call 10 people every day. He didn’t try to pitch anyone. He simply called to ask what they were working on. That’s how Barry stayed top-of-mind and formed business allies.
“If you can’t take the time to learn about your customer and reach out to them on a regular basis, you probably shouldn’t be in business,” Peter says.
7. Treat customers better
People have grown accustomed to frustration when dealing with brands. If you can give them a slightly better experience than they expect, you win.
“You don’t have to be awesome. You have to be better than everyone else,” Peter says.
If you treat them exceptionally, you will create “zombie loyalists,” customers that will drag other people to you.
In order to treat people better than they expect, it requires empowering every employee in your company to create a standout brand. It takes only one to kill it.
If dedicating that much time to individuals seems like it would hurt your brand, consider Peter’s advice.
“You will get the customers you want, by being awesome to the customers you have,” he says.
Students Who Can’t Fail
As a data scientist at Western Governors University, Matt Poulton aims to identify students at risk of failing or dropping out of the online program by analyzing their behavior.
This represents a drastic change in mindset from only a few years ago. For a long time, people did not have confidence in intelligent analytics, Matt says.
“Data scientists get the details,” he says. “Often times it helps you to offer a wonderful customer experience.”
Here are some ways that WGU uses predictive analytics.
1. Barriers to entry
Predictive analytics can let you know the likelihood of someone converting based on their passed behavior.
“Not all leads are equally valuable,” Matt says. “Some bring in a lot of money, some shouldn’t enter the system.”
Just because the leads don’t enter the system doesn’t mean you throw them away. At WGU, they identify people who are likely to fall out of the program or fail and give them extra training.
Providing extra support or content to leads or existing customers could do the same for your business.
2. Recycle leads
If someone is not qualified today, that doesn’t mean that they won’t be at some point in the near future.
“Sometimes it’s just not the right timing,” Matt says. “When you bring in predictive analytics, you’ll know who to go after first (when you) recycle leads.”
3. Track everything
WGU students use portals that track every action they take and capture a snapshot of their screen every 10 seconds.
“We take that data and we analyze as an aggregate how the university navigates the system,” Matt says.
From this data, you can segment it in a variety of ways, revealing homogenous groups that may show groups at risk of failing.
Predictive analytics isn’t a one and done process. You constantly add more intelligence and data sources so that you can improve.
4. Look outside your system
Your customers and prospects interact with more than just your brand. WGU scours the Web to find everything students talk about and funnel and analyze the data.
In doing this, WGU found out that a lack of interaction with people created a motivation problem that kept some students from finishing. The students wanted a support group, so the university provided one.
5. Sell leadership
For predictive analytics to truly have an impact, everyone in the organization has to be sold on using it. That starts with the leadership and includes the day-to-day workers.
Want to celebrate Measurement Week by learning more? Check out yesterday’s blog post here and come back tomorrow for another recap!