Hewlett Packard's Paul Horstmeier has what many business-to-business marketers would consider a "dream job."
While you are getting lead gen campaigns out the door, approving print campaigns, and dealing with your booth for the next big trade show, Horstmeier and his team focus 100% on communicating with customers online.
As Manager of the North American eMarketing team, HP Business Customer Organization, HP veteran Horstmeier has spent the past three years gathering opt-in emails, testing email newsletters, surveying online customers, and revamping Web sites. Here are some of the invaluable lessons he has learned:
QUESTION: What is most effective email name acquisition strategy you have initiated, and why does it work?
HORSTMEIER: Email name acquisition is all about relevance, usually contextual relevance. You need to pick something your customer wants and engage them when they want it.
Within HP, the number one thing people want on our sites is drivers [for printers and other HP products]. If you are a customer and you go and try to find [drivers] and eventually figure it out, then I prompt you at the end of that process to say. ĎI can send you update alerts at any time.í customers will usually opt in. Timing is key.
Customers are sensitive to pop-up windows; so, we try to make the offer right before you look for a driver as well as a ďnotify meĒ box after you have found the driver.
The biggest issue customers have with signing up for anything is spam. Customers do trust HP, but they will not give you their email address until they know exactly what they are going to get. One way we let them know is through a prompt [to receive update alerts via email] after a search experience.
Another way we establish trust is in our newsletter signup. When you register to receive information from HP, we give you a detailed description of what you sign up for as well as a preview of the newsletter.
We test and retest all the time. We used to have a six-page signup, but we found that customers did not want to go through that. We used to want information right up front, such as email addresses and company size and we would get to the content categories later. We found it was better to first let them know what they are going to get, then get their information.
It is a natural consequence. If I ask you for email first, I have not built the trust.
QUESTION: Have you tried sweepstakes or contests to grow your opt-in email database?
HORSTMEIER: We used sweepstakes for signup offers. Anyone that signed up to receive a catalog, complete a survey, or get the newsletter was entered into a drawing to win something.
We found what works best is something from HP like a digital camera or handheld device -- something that speaks to why they came to HP in the first place. We once had a contest where people could win a car, and that was a poor match. We got a lot of clicks; but, we looked at who signed up and it was just poor. Most of them did not qualify for our customer base.
We also found that it is usually not good to offer anything more than $1,000. If the offer is too big, they will assume the chance of winning is low.
Have it be something they want, and create enough chances to win that they care about it. It needs to be personal, something people can use. If we offered a server, customers may not know how to deal with that. If it is a digital camera or PDA, they can identify with it, use it, and it is something that is expected from HP.
We have also tried the Ďeveryone winsí approach and that is very unpredictable. For example, we tried offering Amazon gift certificates [for anyone that registered for an HP offer]. We had one person try to sign up 3,000 times to get the coupons. That did not work. It had nothing to do with the value proposition.
QUESTION: What are some best practice tips on making changes to email newsletters as well as what should be avoided?
HORSTMEIER: There are many different facets to making improvements. One is your email engine, another is design, and another is content.
It was easy for us to think that the way to improve our newsletter was through the engine. That is expensive way to do it. A lot of what you need to improve has to do with the content, it also has a lot has to do with design. When you are not getting the clicks or open rates you need there is plenty that you can do. For one thing you can look at how you label links, where you put them, add variety. These are things you can do without a lot of money.
The best thing to avoid is making quick, knee-jerk reactions to your engine. If you donít get results, donít assume itís the engine. Donít always blame the vendor; it could be the content. Test the content, test small things. Test all the time. Look at all variables.
QUESTION: What have you tried in email communications that has not worked?
HORSTMEIER: Sending out content that is not really credible from HP. A customer might be interested in a digital TV, but they may not expect that from us.
Recently we tried something else that did not work. All I can say about it is the message we were asked to send out was very important to the group that wanted us to send it, and there was no check conducted regarding its importance to the customer. What happened was customers not only did not click, but the unsubscribe rate went up. If you deliver non-relevant information, those that receive it will ignore you or opt-out completely.
QUESTION: We noticed you have recently selected a new email vendor. What advice can you offer regarding the selection and use of email engines?
HORSTMEIER: It was really important that we look long term at our customer base and what our objectives were, regardless of how we chose to segment. We needed to look at how flexible the engine can be. We have two types of customers that are fundamentally different and we needed two unique engines. These engines are able to adapt to our changing needs. We developed these engines with the vendor.
It is really important to identify what your email marketing will achieve. Prioritize your needs. You can over design an engine. If your customers are not going to meet the level of sophistication you can access, do not go there. You can get very wide eyes when you see what you can do. Be realistic. Understand what value proposition you really can deliver.
A typical rule of thumb is most engines will be more powerful than you need. Iím not saying donít use them, but you donít need to use it all. Start simple. Quick and fast is better than long and slow. Getting something going always works better in the ďeĒ space. You have to consider the long term. If what you are doing sticks then you can evolve, but know what you can do right away. Chose a vendor that can evolve with you.
QUESTION: What have you found to be the most important aspects of a successful customer-centric website?
HORSTMEIER: Customers donít want to click on something that is an unknown. Most customers already come to the Web with a task in mind. You have to understand why they are going to come to that site and how can you can complete their task as quickly as possible. Customer-centric sites can go awry if you are not meeting the customerís needs.
However, you should be business oriented. You have to show how you understand the customerís needs, but turn it into revenue. On every one of HPís sites, no matter what, it has to drive revenue and meet the completion needs of the customer.
You also have to adopt a launch, learn, and change mentality. We will launch a site quickly, study it, and make changes. We make changes weekly.
Every site we have ever done when we assume we know the customer we arenít always right. Our objective is to learn. When we do, then after they come to the site we can modify it based on what we learn.
QUESTION: How often do you conduct surveys of your website visitors and how do you go about it?
HORSTMEIER: We are currently trying to do a pop-up survey to grab data on a monthly basis.
We try to collect industry standard data, such as how we are performing. We ask questions on satisfaction, how likely are you to return, how quickly can you find info.
We do this every month now, but in the past it was every few months. At that rate we were not finding out enough, and we were doing it in a very proprietary fashion.
We have seen that we can get pretty good responses as long as we keep questions to just a few. Customers will definitely be more receptive this way. If you want to ask more questions, you can vary the kinds of survey questions they will see. You can also cookie to know who has seen a survey so they donít have to see one again for a while.
QUESTION: How do you measure the success of your loyalty programs?
HORSTMEIER: I have four sets of metrics. One is activity and effectiveness, which includes things like total emails delivered, total opens.
The second is total customer experience; this is for surveys when we find out how satisfied were you, if you would make recommendations and if you would come back. I do that on a scale of 1 to 5 and within that I have a loyalty factor. The higher the number the more loyal they are.
Revenue is third. I measure that on email, on the Website, and leads generated.
Fourth is cost effectiveness. I measure how much I save by reducing support costs. My team works with the customer support call line.
We also measure the amount of marketing savings we have, track the cost per touch on Web and email, and compare them to our traditional channels.
QUESTION: How often do you conduct surveys of your email opt-ins, and how do you go about it?
HORSTMEIER: We have a quarterly goal, and there are three types of surveys. One is for satisfaction and loyalty where we ask thinks like if would recommend this to friend or if you were likely to use this again. Another contains profiling questions where we will ask customers to tell us about themselves and their company. We also have one where we ask about content they are getting in their emails as well as what they would recommend in the future.
These surveys are sent within the email newsletters they already get. We will highlight a link and ask them to take a few minutes to it fill out. We will also offer them the opportunity to get a digital camera or some kind of offer for completing survey.
We have done for this past year and a half. We only survey sections of our database at a time [rather than all at once] and we keep track of who clicks. If you click and fill out survey, you will not see a survey again for a while.
QUESTION: In a survey you conducted last year, 50 percent of people who received HPís email newsletters gave the newsletters the highest possible rating on your three key loyalty questions. What did you learn about your readers?
HORSTMEIER: We have learned using websites to capture email names is the most valuable way to get the most loyal customer. Email is a good way to get in touch with your install base.
We think theyíre happy with email as a medium if we consolidate our messaging into one vehicle. They love getting one monthly update rather than five or six from different product divisions.
We learned click through rates are not necessarily an indication of success or disinterest. Users may look at an update and think itís great, but they donít need anything that month. At least they were updated and thatís most important.
NOTE: Would you like to meet Horstmeier in person? He is presenting at ad:tech Los Angeles on June 19-21st at the Westin Bonaventure Hotel.
For tickets and travel: http://www.ad-tech.com