MarketingSherpa Video Archive

Optimization: A discussion about an e-commerce company's 500% sales increase

Daniel Burstein, MECLABS, and Steve Parker, firstSTREET

In this MarketingSherpa webinar, we spoke with Steve Parker, Vice President, Direct Marketing, firstSTREET.

Learn how his one-man team used a radical redesign to achieve radical results, thanks to the power of compounding multiple wins over time. See how Parker achieved a 3,566% increase over the control page right away, and how he continued on to a 500% sales increase.

You will gain insights into how to understand your customer's thinking, which for Parker, led to a 46% conversion rate increase on a homepage.

In this webinar replay, you will learn:

  • How Parker created a testing and optimization program with a "three-person team" — as he put it, "me, myself and I."

  • How the program began with minimal resources — a bare budget, not a lot of time that could be dedicated to it and with Parker having no coding or graphics skills.

  • The best way to begin a testing and optimization program is simply to get started — take something simple on a landing page, run a test and continue from there.

  • How a radical landing page redesign created radical results — a 3,566% lift in conversion.

  • And finally, how additional optimization on the new landing page resulted in a 500% increase in sales, directly impacting bottom line.

Download the slides to this presentation



Related Resources

Landing Page Optimization: Radical redesign leads to 3,566% increase in conversion

The WOW! Computer: How firstSTREET applied a radical redesign to deliver a 500% sales increase

Sierra Club Tests Radical Email Newsletter Redesign to Improve Preview Pane Viewing

Testing: Go big, or go home?






Video Transcription

Burstein: Hello, and welcome to a MarketingSherpa webinar. Thank you for joining us today. Today, we're going to be talking about optimization. We've got a lot of questions from our audience we're going to answer today, and, hopefully, we'll be able to provide you an inspirational story to help you improve your own optimization efforts.

My name is Daniel Burstein. I'm the Director of Editorial Content here at MECLABS. I am in our green screen Studio right here in Jacksonville Beach, Fla. Joining me today is Steve Parker, the Vice President of the Direct Marketing Division of firstSTREET. He's in Richmond, Virginia right now. We thank you for joining us, Steve.

Parker: My pleasure, Daniel. Good to be with you.

Burstein: I'm going to be posing a lot of questions to Steve about his very impressive case study, in which he increased sales. Well, we're going to talk about that in just one second. I don't want to give it away.

First, let me let you know you can use #SherpaWebinar to ask us any questions. Put some questions on there that you'd like me to ask Steve. You can also share your own advice about what works for you in optimization. We found a Twitter hashtag is the best way not just to communicate with us, ask us questions, show us what works for you, but it's a great way to find your peers, too. Find some optimization peers you can work with to improve your own testing efforts.

Also, I just want to throw these up here real quick, so we have them for the replay. We're going to put this out through YouTube, we're going to put this on the MarketingSherpa site on SlideShare, and there's some further resources you can look at to learn more about Steve Parker's story. How he did what he did, some of these impressive results. I'll also be tweeting some of these further resources through Twitter using the #sherpawebinar.

With that, let's start with the fireworks, shall we, Steve? Let's talk about what were the results from your efforts?

Parker: In 2011, I attended Optimization Summit as one of the students in the room. I was new to the concept, but it seemed that we were pretty good at driving traffic from a lot of different methods, but you know what? We had some websites within our portfolio that really weren't doing very much with that traffic. I came in, I listened, I tried to apply what I learned, and, on the study we're going to talk about today with the “WOW computer,” we literally increased sales a little over five times. And that's selling a very simple computer to consumers who, many times, aren't even online. Which was kind of an interesting challenge to begin with, so we're almost selling it indirectly through their kids. Our target market being senior citizens and the elderly.

Burstein: Yeah, that's very impressive. That's something to think about. That wasn't an increase in just clicks or email opens or anything, that's an increase in sales. That is very impressive, Steve.

Now that we've got an idea of the results, let's get into the back story of you as a person, as a marketer. I always say "the past is prologue," and when I was looking into some of your background on LinkedIn, I found that you had an interesting background. You worked both at consumer products goods companies like Keebler and the company behind No Nonsense hosiery, but you were also involved in really the first incarnation of the internet, working at companies like MyPoints. I wanted to get a sense of how did those previous jobs inform your current work at firstSTREET?

Parker: Both of them, whether it's consumer packaged goods or MyPoints being a direct marketing model on the Internet, they are heavily quantitative. Heavily analytical. Not that we've got a bunch of statisticians running around here with Ph.D.s, but the idea of looking at a marketing funnel, for example, that was something I was doing back when I had to do it in Lotus 123, pre-Excel days. I think I just dated myself.

The idea of just understanding what changes have what effect financially on what you're doing, that's a Core 2 Optimization, is something that had been baked into my career. What I heard about this concept that MECLABS was teaching about how to really optimize webpages, it just intuitively made sense to me, and I wanted to learn more.

Burstein: Let's get a sense now of your current role at firstSTREET. Tell me a little bit about your current role, and a little bit about what firstSTREET does, so we can understand the test we're about to see.

Parker: Sure. firstSTREET is the largest direct marketer to senior citizens and the elderly in the United States. We sell products that try and make their lives better, help them offset some of the challenges that come with age, as eyes, or ears, or mobility decline as we get older. My responsibilities include our flagship catalogue, our internet businesses, the new product development and merchandising pieces of the company, inventory, some business development, and a couple of other little things.

Burstein: Our last webinar was with SAP, and they were telling us about their testing efforts. Obviously, SAP, I think in the webinar we learned, has 60,000 employees all across the world. A lot of questions we got from the audience were, "Well, that's great. SAP can do it. They've got all the resources in the world. What can I do? I'm just a small company." So, Steve, give us an idea of what are your resources in your current job?

Parker: I'm pretty much the opposite of SAP. At the time we did this WOW computer work, we had a three-person team. That team was me, myself and I. Since then, with some success, the team has grown. I have one other person that is a tremendous asset in the marketing that we do, but we had to approach this bare-budget, minimal time, minimal dollars to be invested against it, and I'm not a coder. I'm not a graphics guy. I'm a marketer by training, but I don't have the Web production skills, so I even had to find a testing tool that was, basically, WYSIWYG, or drag and drop, whatever you want to call it, where I could move pieces around and change text, without having to involve IT, without having to involve anybody, really.

Burstein: That's got to be very challenging because you're doing everything yourself. On the up side, I would think, maybe it's a bit easier to sell it to management if it's a smaller company with less resources. That's another challenge we get. How can I sell it to management? How can I get the resources I need? What was your experience with selling it to management, if you wouldn't mind telling us?

Parker: Initially, it was met with a little bit of skepticism. Is it the best use of your time? If you've got X-number of hours a week you're going to spend on all the things required for internet marketing, where are you going to get the biggest bang for your time? Biggest bang for your buck, so to speak.

I started small. I took a website for a product that we didn't have great hopes for online. It was a safe place to test. That helped a lot with not having many real issues around are you going to damage the main business here.

It was something that I basically did on the side, so to speak. A lot of Friday afternoon work on the testing when things were kind of winding down. For me, the tests had to be relatively simple. I'm not doing multi-variable testing. I'm not doing lots of fancy graphing treatments. I'm not doing pieces of technology. Much more focused on simple communication to the consumer. Are we telling them what we ought to be telling them, and in the right order?

Burstein: I understand you had an important meeting this morning. I don't know if you can talk to this at all, but I wonder if there is anything from that meeting you can share about what management now things about your testing efforts, that you've seen the success.

Parker: Yes, we had a board meeting this morning. Testing is, as direct marketers, something you always want to be doing continuously. We have evolved from my little side project to quite a bit of testing. We've even moved as far as doing some online testing as a precursor to offline print ad testing where I can test cheap and fast relative to testing an offline direct response print ad. So, it's trying to make us a little smarter, and help us be a little faster in finding the good ideas from the bad, and then being able to execute them on a bigger stage.

Burstein: Clearly, Steve committed big-time to testing, taking it on himself. Now, it's time for you to commit, in the audience. This was an idea Steve gave me. I think it's a great idea. A lot of times, we listen to webinars, it's easy to just kind of listen in the background. Tune it out, and not really get full value out of it. This is our challenge to you. Me and Steve are making this challenge.

Write down the one best thing you learn in the next 30 minutes. Something that you really want to act on. That you'll commit to acting on in the next five business days to put this learning actually into action. That's essentially what Steve did. He said he attended Optimization Summit 2011, he learned something, he went back, he put it into action.

Now, from his experience, I know he's had experience before teaching, and he gave us this idea. It's not enough to say I'm going to do this, you have to commit. You actually have to write it down. I challenge you to write that down sometime during this webinar. Use #sherpawebinar, write it down on Twitter, send it out to the world, send it out to your peers on sherpawebinar, saying that you are committed, you are motivated, and you are going to make this change to improve your marketing efforts. That's our challenge. Let's see if you can live up to it.

With that, let's jump a little more into this case study. So, you came back from Optimization Summit 2011, you took a look at this page, and, for those who aren't familiar, at the top there that C=4m kind of formula-looking thing, that's called a MECLABS Conversion Heuristic. You can search MECLABS Conversion Heuristic. You can learn a lot more about it. It's basically a way to break down the likelihood of conversion on your landing page to help you optimize your landing page.

As you look at this page from Steve, wow. He sees all these different changes that he has to make, and the first thing I think about is a question that we got from Matthew. He wanted to know, "How do I get other Web designers excited about testing, and how it can make them better designers?"

When I look at this page, Steve, this just looks like a designer's nightmare. How did you get them excited about designing a page like this? I know it was mostly you, but I think you worked with some freelancers, some other designers?

Parker: I did, yes. Once we had some initial success, I did bring in freelancer design help to get us to the next level with the re-design. The main thing for a designer to understand is we're not, on the Web here, in the business of creating pretty art. If you're looking to end up in MOMA someplace, to be the next Andy Warhol, that's great, but we're looking for designs that drive results.

In my case, I have a very old target market. They read. They have that kind of time. They are going to spend more time on a webpage than a 20-something or 30-something is going to do, and they're used to an advertisement style. They've seen kind of what we would call "old school, heavy text, ugly stuff." They've seen that for decades, and so I'm trying to put something in front of them that is familiar. It's not intimidating. There aren't things moving around when they mouse over something. They don't realize that's common on the Web, necessarily. It's a design that's based on "We win on financial results. We don't win on pretty pictures."

Burstein: As we look at this page, we see that there's a lot of changes you wanted to make when you got back and you looked at your landing page. We had a question here from Peter. I think it ties into it very well. The steps you took. What steps should be taken when my site has been somewhat abandoned for a time?

Parker: That's kind of what we had here. Rule number one is get started. Do something. Pick a safe test. Maybe a test doesn't have to be 50-50 if you have a lot of traffic to a page. Start with a smaller piece so you're not taking a huge risk, and look for the lowest-hanging fruit. You get a win under your belt quickly; it gives you some momentum to then try things a little bit bigger. All the arrows on the slide show you that we tried a bunch of stuff. If it's red, it failed. If it's green, it was good, and the thicker the arrow, the bigger the win or the loss.

We came and beat this page up pretty hard over the course of several months, actually, and there's some green up there, but there's an awful lot of red and black, black being a tie. So it’s get started, and keep iterating from there.

Burstein: I think that's a great point, Steve. I love how you have that big red "oops" on there. It's OK to admit when you're testing, sometimes you're going to be wrong. One of my favorite things you said when we were preparing for this webinar was, "It's OK to fail a lot along the way." Keep that in mind if you're just starting optimizing, testing. You're not always going to be right, but one thing, the great thing about it, when you're wrong, is you learn something.

And so, I've got this question here on #SherpaWebinar from Grant Baker, "Where did Steve gain his insights into his market?" I don't want to give away too much about your optimization summit session at Boston next week, but I thought it was really interesting how you learned about your market from testing. Can you tell us just a little briefly about that?

Parker: Sure. There's lot of ways you can understand your consumer. Direct consumer research, or primary research, if you have the budget and the time, great thing to do. Secondary research, meaning what's already out there, published on the internet or offline. And, from an experimental standpoint online, we tried different "positionings," if you will. The sub-headline. We tried two or three different types. We found one that stuck out better.

Next week in Boston, at the Optimization Summit, I'll be talking about Jacuzzi walk-in tubs, a partnership of ours. We found a different positioning through testing, and we're running with it across the company print offline. We're even wrapping bands now that have it all over it. If you're going to stick it on a band, it's pretty permanent. You're committed to it at this point.

Burstein: Excellent. Let's jump in to one of your controls. The interesting about this control that I notice is I don't see something that I see on many landing pages, and that is some sort of social media sharing button. We had this question here from Ilene [SP]. She wants to know, "Did you combine website improvement testing with social media or content marketing?"

Parker: In our case, given our target market, you're looking at really an age 75-plus customer, they're not big social users. Of the ones who are on social media, they really just want to see their grandkids and their kids, so they're not going to be as interactive in the social world. From our standpoint, it's pretty low on the priority list. There are no social buttons on this website. We've tested a little bit of that on some of our other properties. As the baby boomers, the younger part who grew up with some social media in their lives, get older, yes, it will definitely be more important. For my particular target market, at this point in time, it doesn't help.

Burstein: I think that's a great point that we can all learn from, Steve. While we might not all have senior citizens we're targeting who are more averse to the social media, perhaps, even if we're targeting a younger audience, it's finding which platform they're on, right? There's no one right platform. Some people might be on facebook, Pinterest, LinkedIn, Twitter. If you're using social media, it's definitely about understanding your audience, not just a fad, and understanding what works best for them.

With that, let's take a look at this control. Something I really noticed about the control was that button. That's a very interesting button, and we had a question here from Matt. He says, "I'd like to hear about call-to-action buttons, color and size of the button, font color and size of the text on the button. Thanks." So, can you give us your thoughts around buttons?

Parker: We actually did a button-only test, and, with minimal effort, got a 15% lift as part of this re-design. The way our buttons are designed, particularly on what I call "high friction, high anxiety" clicks where you're asking for an order, you're asking for a lead to be submitted. We were looking to bring more to the value.

MECLABS jokes about submit buttons, and you will submit to the marketer. I think everybody who's been on more than one of your webinars knows better than that now, but just a button that says "Add to Cart," it's functional. It doesn't convey a benefit. It doesn't help me feel safe, or secure, or this is a good thing, or reinforce anything that I've done. So, we went for a button that has some communication value to it. It's a little blurry, but "Get your WOW computer." So, get. As opposed to "do something," it's like "get something." Below that, "Easy, Safe, Guaranteed."

We're trying to minimize the anxiety of, you know, this is a reasonably expensive computer, and we're asking you to buy it, so I want to make you feel as safe a s possible that it's going to work for you, that we're going to stand behind it 100%, and every little bit of push I can get towards getting somebody through the funnel to make a purchase increases the response rate, and we all make more money.

Burstein: I wanted to ask briefly, too, I noticed this is a pretty long landing page, you have the button both at the top and the bottom of the page. Do you have any thoughts around button placement?

Parker: Yeah. More than one, on a long page, certainly makes sense to us. If you think about traditional selling, you'll hear people talk about a trial close. Webpages don't sell product, it's the people behind them. It's the story, it's the "what are you buying into." So, if there's an appropriate place to check and see if you've told someone enough to get them in, we could potentially put more than one button after that first one in a design like this, because this is a pretty long page. Just try and have a conversation with your customer. You're doing it with a webpage, granted, but what would be the logical thing to say at each point as you run through the benefits, and then the features, of why you've got the greatest thing since sliced bread that these people should buy immediately.

Burstein: Let's take a look at these two pages side by side, this re-design. We have a question here from Margaret. This is a common question we get. People re-design their pages, and they're worried about the all-powerful Google. "I don't want to mess anything up. I have a decent Google page rank. If I change anything, I could lose it. Even if, by changing something, I can increase conversion."

You might not have had as big of a challenge here, because you went from a shorter page to a longer page, and people tend to be worried about cutting tons of copy off a page even if all that copy inhibits conversion. Nevertheless, Margaret wanted to know, "How did the re-design effect Google rankings, back links, etc., and if not, what steps were taken to minimize negative ranking effects?"

Parker: Let's start with the technical side of it. Back link-wise, we didn't really change URLs that were in the original mini-site there, so we kept the same URL. We changed the content on it. You can certainly use a 301 redirect to keep most of the links flowing if you have to re-navigate your site as part of the change. From the work that we did, yeah, we saw a short-term dip in Google, but the vast majority of my traffic is coming from offline print advertising, and from the catalogues that we send out. We're not doing a lot of paper click. We don't rely on organic for a great big piece of the traffic here, so it was less of a concern for me than it sounds like it would be for her.

The other thing that we did as we went into this is just made sure that we had our SEO basics correct. The question about someone whose site has been static for a long time. Well, Google's algorithm has not been static for a long time, and what was a brilliant idea in 2010 is probably working against you in some ways now. Whether it's links, how you're handling your metadata, or not having metadata, importance of titles, importance of URLs that are the same as your T-word, the game is rewriting itself pretty quickly. So just by cleaning up and catching up, if you will, with more current SEO best practices, we actually came out better on the back side of it in fairly short order.

Burstein: All right. Thanks. Let's take a look at the results now. Wow, that's 3,566% increase. That is not a typo. I've got to say, Steve, I've never had a lift that big. If I did, I think I would make a T-shirt with it, and I think I would plaster it on the walls of the office so everyone here knew about that lift. With that said, we get this question often, and we got it today. How do you organize the results and keep track of them? Short of making a T-shirt with your results, how do you keep track of your test results?

Parker: The first part of that is if your lift is that big, what it really tells you is how bad your previous page was, and the question you should get from your CEO is, "Why haven't you fixed that already?" That's just too big a lift to expect to see.

In terms of how we handled the results, we do a decent number of tests, and it's not something that I think we're particularly great at. I've never seen a really good third party solution that creates the perfect database with everything you'd want to have in it. Certainly, some of the services we use, email marketing, for example, the service we use for our A/B testing, they have somewhat of a repository of test results that you can keep there, but they're not even perfect.

For example, keeping good copies of what your control and your test were at the time you did them, even in the offline world, if somebody misplaces a pile of direct mail samples, you may have lost your test and control reference file if you don't have a good wraparound on it. That's one I'd actually ask you about, Daniel. MECLABS orders magnitudes more tests that I do and you're way more sophisticated than I am, but I would think it's even a challenge for you guys given the breadth and the volume of the tests you've done over the years.

Burstein: Absolutely, and, really, it's another data point. As will all marketers, we see data as a challenge. What we've done here is we have a very rigorous process of having test protocols, so in MECLABS, each test is assigned a specific test protocol. That includes all the important aspects of the test, which includes a hypothesis, why the test was run in the first place, the different designs and treatments and controls, and the results from it. In that way, we can go back and look at previous tests, and distill learnings from each one. In the end, why we're testing, why we're hoping to test, all of us, is we want to improve our customer theory, right? It's not about one page, it's not about knowing one button works without the other, it's about really learning about the customers.

Like Steve was talking about previously, about how he learned about his Jacuzzi customers.

So, there's no one right answer that test protocol works for us. In the end, the technology is pretty simple. It's essentially Excel documents and a Microsoft SharePoint, but just finding some sort of rigorous process to keep your tests, and to communicate them. As we learned when we talked to SAP in our last webinar, one of the most important things they did, I think, in such a big organization, really any size organization where there's more than one person, you have to do this, they communicated those learnings. They kept a rigorous account of what they learned, and then every quarter, they would communicate that out to the entire company.

Also, this helped when they got test requests. For example, a test for a landing page, a test for a button, some of the things they felt comfortable saying, "We don't need to test this. We've tested it enough, and we know what's going to work best. We can give you, at least, not know for sure, but at least a best guess of what you should do with this button. That way, it saves some of the testing bandwidth.

One more question I had for you, Steve, about this. As you said, that's a big lift, and, very humbly, because maybe the initial page underperformed, but Terry wants to know, "What was your most important website change?"

Parker: Really, the most important one was deciding to do the radical re-design itself. You saw, in the first slide with all the arrows, we beat up the existing page pretty hard. At some point, it finally got through my thick head that, you know what, this page just isn't very good, and no amount of tweaks and twists and turns is really going to make this thing go. So, it was making the leap to, "All right, let's just start over." Brand new idea. We brought in a new outside freelance designer so we'd have a little fresher thinking. Wish I'd done it six months earlier. Hindsight is 20/20, but I would say that, by far, that was the big change.

Burstein: Let's quickly look at some follow-up tests you ran. Here we go, is an A/B/C test. I believe this test was about price. How should you communicate price? What did you learn about communicating price?

Parker: Two parts to it. In our print ads and our catalogues, for this item, we don't put the price in the ads. Our experience is we will, at the end of the day, sell more units of whatever product it is, on our major media items, if we don't give you price until you call. What it really does is it gives us a chance on a one-to-one basis on the phone to build up the value proposition. The more you understand how it fits your needs, the more that it's something that would really be relevant, that would make the quality of your life better as a senior, the better chance I have that price really isn't an issue anymore.

So, what we tested here was giving you the price up front, at the top of the screen there, or telling you that "call now for your best price," or just not even addressing the issue and leaving it blank. Control was B, on the left, which was "call now for your best price," and that won in a landslide, so we saw the same type of behavior online that we typically see offline.

Burstein: So, price is a question we often get. Another question we often get is about content length, and you ran a test about that, too. We can see the control, variation one, variation two, and in this one, the middle length one was up 46%. Did that change how you approach content length?

Parker: It did change how we did content length. What really came out of that, 46% gain is awesome, but a better understanding of what's most important and what's not as important to our consumer. I think that was the real learning out of that. In the shortest version, we cut too much out. We didn't tell them enough of the story to really establish the value proposition, but in the long one, we were running on a bit at the mouth from a marketing standpoint. So, we learned a couple of the benefits were nice, but in thought sequence, in order of "how do I get a consumer's attention and then keep them interested? Draw them deeper into how this product fits with their life," we had some stuff that was taking people off track. By basically "chunking" the site, and by looking at what we could pull out, we learned what the best mix would be.

Burstein: Again, that's a great example of using testing to learn about your customers. Here's a very interesting example in that you're driving traffic from print ads to the Web. You can see on the left and the right, the big difference is that very clear headline. I believe that was a headline from the print ad. "Wow. A computer designed for you, not for your grandchildren."

OK. Don't tell us what won just yet. I'll ask you that in just a moment, but first I want to tell you about Optimization 2013. It is next week in Boston. We hope to see you there. You can go to, O-P summit, to learn more about it. Steve will be there. He will be speaking about what he learned from his customers and how he learned from his customers by testing.

There's a very interesting story that we were talking about previously about trying to figure out what resonates most with your customers, and how testing can inform that, and how sometimes you have to realize you don't only have one customer. So, finding out what resonates with different customer segments. I'll be there, too, and I hope to see you there next week in Boston.

But, let's take a look at these results. Forty-eight percent increase in engagement. I think the biggest thing to learn from that is to make sure that there's a consistency in thought between the ad, in this case it's a print ad, but this could apply to a PPC ad or any other type of ad, and the landing page. What do you think about that, Steve?

Parker: Absolutely. And, again, this is something I learned in 2011 at the Optimization Summit, but if you attend the MECLABS webinars, you've heard Dr. McGlaughlin and others say the first thing your site has to do is, "Where am I? What can I do here? Why should I do it?" Well, if I drove somebody from a paper click ad to a landing page that's along the same theme, you're trying to address those things. In this case, I'm driving you from Parade Magazine, or AARP Bulletin, or some magazine, or, occasionally, our catalogue to this website.

So, the leap is even across media types, and by having the same headline on the website that you saw in the print ad, we got a very nice lift simply by answering the "where am I," and letting the customer know that they got to the right place. Again, I've got older customers. They're not as sophisticated internet users as a younger demographic would be, so the, "Where am I," and "Am I even in the right place," is probably even a bigger issue and a bigger opportunity, if you can bridge that gap for them.

Burstein: You have a very interesting funnel in that you're primarily interested in driving people from offline, traditional advertising, such as print advertising, to a landing page. Do you have any special considerations you take because of that?

Parker: We do. From a design standpoint, we have a bigger type size than most people would use. We have to have very high contrast because, let's face it, the eyes have faded out and our capabilities over the years. We don't want things that are confusing. We try and keep it as simple as we can, and we're trying to have a conversation with a customer that has seen decades and decades and decades of marketing. We're not quite as low on the opinion scale as a used car salesman in a senior's eyes, but they don't trust marketers particularly, so we need to be erring as much as possible on the side of minimizing any friction, minimizing any anxiety.

They're looking for a square deal. That's what we try and provide, and we have to do that from a design standpoint, from a communication standpoint, and just from having a conversation. They're used to an environment where people talked to you and sold to you. Where you went into a store and there was actually help there that even knew the products, so to the extent that we can replicate that online, or in an offline print environment, we find that we have more success.

Burstein: Clearly communicate, provide, and sell value. I love it. When we talked about some of your initial results, Steve, you were very humble. You said if you get an over 3,000% lift, the CEO should call you into the office and say, "What were you doing wrong before?" The CEO is not here right now. You're among friends. Brag a little. Tell us about this compounding success of yours.

Parker: We had very low expectations for the WOW computer on the internet. You're trying to sell computers to people that don't have a computer in some cases. You've certainly got an influencer involved in their kids. If Grandma wants to buy a computer, she's going to call somebody younger to have them come check it out. What we found is that there's a lot of folks out there using computers that are too complex. They don't understand how they work as well. They just aren't having a good experience online, and if you think about it, online is a great way to keep a senior connected with the world when maybe mobility isn't what it used to be.

What I think we learned out of a year of this is the value of compounding. It's a lot of little wins that add up to a huge increase. There's a bunch of failures in here I didn't put in. There's one no-gain, but we didn't bat 1,000. We didn't even bat 500 in terms of winning tests, and I wouldn't expect to, but the things you test that don't work, they go away. The wins that you get, they compound, and 30% win, 100% win, they're nice if you add them, but when they start building on each other, and if you do this for years, you really see some pretty exponential growth in your results. Persistence, I think, is critical to it.

From my own perspective, what that kind of work got me was the ability to hire a new person who, fortunately, he's a stud, and a lot more ability to test because it's affecting our bottom line. We're helping people get products that make sense to them, and we're making more money in the process. That, to me, is a win-win all the way around.

Burstein: Excellent. We have just one minute left, Steve, but if you could sum up your top three lessons to other marketers so they could replicate your success, what would you tell them?

Parker: Number one, I'll steal a line from Nike. "Just do it." You've got to get in action. This is why Daniel challenged everybody on the call this afternoon. Take some action. Get started. You can certainly fail along the way, and, quite frankly, your failings should teach you something.

It's either a learning experience, or it's an earning experience. Occasionally, both, but if you fail, and you can pre-plan this, if you put a test out there, if it wins, great, if it ties, OK, if it fails, what did you learn because of the failure? So, you can do better with that information the next time.

The third one, calling it "tribal," working with your peers. This is an environment where, whether it's "crowd-sourcing" is a buzzword you like, or a mastermind group, or whatever your way of networking with friends, getting input from other folks who are on this same quest of knowledge and improved results would be, the results will be much better if you get more input brought into the process. You still have to run it, you still have to measure it, but trying to do 100% of yourself, like I did, we're doing a whole lot better now that I've got a broader set of people working on this with me.

Burstein: Thank you, Steve. Thank you for joining us today. I look forward to seeing you next week at Optimization Summit in Boston, and thank you all for joining us. Please join us for our next MarketingSherpa webinar where we'll be discussing content marketing. Good luck with your optimization.