Conversations on Conversion interview with Khalid Saleh

Conversations on Conversion is an interview series featuring conversion optimization thought leaders. We started the podcast back in March of this year, and frequently release new interviews. You can download the podcast for free from iTunes.

This episode features a lively discussion with Khalid Saleh. Here’s a transcript.

Anna:  Welcome back everyone to Conversations on Conversion. I am Anna Talerico and with me today is Khalid Saleh. He is the president and co-founder of Invesp, a long-time e-commerce conversion rate optimization company. Khalid is a frequent speaker, so you probably already know him from his many speaking engagements at events like SMX,  eMetrics, PubCon, DMA, and others. He is a sought-after conversion rate expert, especially in the e-commerce arena so we are really excited to have him today. Khalid, welcome.

Khalid:  Hi, Anna how you doing?

Anna:  Great, I am so glad to finally be sitting down with you on the record to talk a little bit about conversion rate optimization, which as you know has had a bit of buzz this year.

Khalid:  Oh, definitely. It is sort of funny, because back in ‘06 when we used to talk to companies about conversion rate optimization, there was not a whole lot of focus on it. But I think with the down economy, most companies are starting to see how can we really turn more visitors into actual customers, so there has been tremendous focus on it since last year, actually.

Anna:  And because of this new focus there are a lot of people just getting started. The question of the day seems to be, how do you get started? What advice do you give to marketers who are just getting started with conversion optimization?

Khalid:  There are a few things I think every marketer should keep in mind whenever they’re starting with conversion optimization. First, I always tell marketers to go for small, consistent wins. Those are a lot more important than one time, inconsistent, big wins. 

Lots of times when we work with a client they’re looking for a tremendous number of changes. They want to change everything on their website. The problem with that approach is you do not know what’s going to work for you and what’s not going to work for you. Really, what’s important to you ultimately is to increase conversion rates. It doesn’t matter whether you change a lot of elements or not. You should really evaluate a change on your website by the results, not necessarily by the amount of work you’re investing. 

I was just talking to a client yesterday and we were working on their checkout process. He was expecting a major change with his checkout, but we just tweaked it in a couple of areas. This gave them about a twelve or thirteen percent uplift, and he came back and said, you know, I expected to have to do lots more. 

What I always tell clients is to not evaluate the size of the recommendation. Evaluate the size of improvement, because that’s a lot more important. 

The other thing is, you should probably spend some time really focusing on the analytics because where you start is extremely important. We work mainly with ecommerce websites. Our typical clients have anywhere between ten thousand to twenty thousand items in their catalog, so you can imagine they’re sometimes about two hundred to three hundred category pages. Then there’s the main home page and there’s the checkout process. The question is where you should start because ultimately you have limited time, limited resources, so you need to really pick the right page. 

Analytics experts, consultants, if you have your own analytics team, you should really spend quite a bit of time looking at the numbers story behind the websites to tell you ‘well, here are the areas of the website that are working really well. Here are the areas not working really well.’ And based on that, you will see where you should start. 

Finally, I also tell clients, if you’re just starting with conversion optimization, in the first month or so when you’re deploying your first test- let’s say, whether it is A-B or multivariate, just go for something small. You know, change the button on a call-to-action or just a tagline. 

Ultimately, the goal of that is to get your technical team familiar with the testing technology that you’re going to be deploying and working with over hopefully long term, long period. 

The first couple of months you’re just learning the technology, making sure that it’s integrated and plays nicely with your own ecommerce technology, or whatever technology you’re using on your website.

The first two or three months we call research months where it’s really about figuring out where you should start, deploying small tests, getting really comfortable. The third or fourth month is when you start aiming. 

Anna:  I love the advice to start small, because I think that’s why some companies never get started or they flail around. When you get started small you eventually look back and end up seeing a big body of work that you’ve accomplished without having to bite off everything at once. 

Khalid: Yeah, it’s a lot better to focus on the long term in conversion optimization. What will pay off as opposed to quick gains, or too many changes.

Anna:  Sure, I think biting off too much is probably one reason why conversion programs go off course. What do you think are the top three reasons that conversion programs go off course?

Khalid:  Whenever I’m talking, I’m talking from the lands of ecommerce. Last year we talked to over two hundred, two hundred fifty different ecommerce companies, some of the insurance retailers, top 500, and the consistent thing that we hear from them is they’ve tried conversion optimization. They’ve invested the resources and it did not succeed. 

At some point, I think about half way through last year almost seventy-five percent of the companies we talked to really were turned off by testing. They tell you ‘well, you know you guys talk about it all the time. Tell us we can achieve results but we’ve tried it and it didn’t succeed.’ One of the reasons for that is really not picking the right type of test, whether it is an A/B or a multivariate test. 

We were just talking to a client of ours recently - it’s a huge website. They have about four million visitors a month and they tried to deploy a test. They got to their checkout process and they said you know what, we’re going to optimize the first step of the checkout process. They went ahead and created thirty thousand different combinations from that checkout process.They were deploying multivariate tests.

You know, the theory was ‘what are the chances that those thirty thousand combinations would be beaten by the original design.’ So the test ran for a couple months and low and behold, the original design beats all those thirty thousand combinations. So, just this random guessing, hoping one of the combinations is going to win sometimes really backfires. 

The same client just deployed an A/B test where they have the original, and we gave them two other versions, and both versions are beating the original. 

The other thing that many clients do not think about is having the resources to implement testing. We see it because we’re a consulting company. The clients are ready, they’re excited about testing, but they have not thought about the amount of time it requires or the technical resources to implement the task. That’s very important because they are two big elements in conversion optimization. 

While there is figuring out where you should start optimizing and what you should change,  the other critical part is actually having to do the work. That becomes an issue because yes, there is excitement about conversion optimization to start with, but that sometimes fizzles out because your IT team and your development team have about a million other things going on, and conversion optimization gets pushed back…especially the first few months. 

The last thing is stopping too quickly. Too many companies try conversion optimization for a couple of months and they don’t see the results and they say you know, we’ve tried it. We’ve had enough. Let’s just move on to the next thing. 

I always tell clients if you’re going to do conversion optimization for less than four months, you probably should not even plan on it. Four months. Give it time. Give it its due process. 

Anna:  I love that actually. I always say at least three months, so I’m going to change mine to at least four months because I think you’re right. You’ve got to start with a minimum commitment and four months sounds like a great one. I’m also glad you mentioned resources because I’ve been doing these Conversations on Conversion for a while and there’s not been a lot of talk about resources, and I almost think it’s a bit of our dirty little secret in the industry. It is a lot of work, whether it’s with technical resources, whether it’s the strategy, whether it’s the analytics, or design. You know there’s so many different types of resources that could be part of the conversion program, and I think you’re absolutely right that not being realistic about the resources is probably one of the biggest downfalls of programs. I’m really glad you brought that up.

Khalid:  Yeah, we always tell our clients ‘If you want to have a successful conversion rate optimization program, understand it’s a collaborative effort.’ There are the conversion optimization consultants or the in-house team, and there is the technical team, and there’s the SEO. Really, everybody has to come together for conversion optimization to succeed.

Anna:  Definitely. So, I know this is a question you get a lot too, and I always hate to ask it because we all know there’s no magic bullet, but I do like to give people tips. Here we are talking about how much work it is to optimize, and this is sort of the flip side of that coin. There are so many little things that we can do to help move the conversion rate forward but if somebody could just do one thing today to help with the conversion rate, what do you think that one thing should be?

Khalid: We’re working on a book for O’Reilly on conversion optimization and one of the metrics that we use, which I think is really helpful and not many people talk about, is what we call visitor abandonment rates.  I’ll talk about it from an ecommerce perspective. 

If you think about an ecommerce website, the first way that you can tell a person is engaged with your website is whether that visitor is adding an item to their cart. So look at the percentage of visitors who are not even doing that simple interaction. 

Let’s say you have a subscription website or let’s say for example you have a website that offers white papers. The very first level is a person maybe filling a form. What is the percentage of people who are not even taking that simple action? It is not necessarily a full conversion. For example, on an ecommerce website that gets fifty thousand visitors where only three thousand are just adding items to their cart, that tells you that there are major issues on the site, whether it is a site design, navigational issues, and product pages. So those are big things that you can focus on and really kind of start the war against low conversion rates with before even going and figuring out the other aides of conversion.

Anna:  Love that. Definitely. So, one final question before we wrap up: what do you think are some trends or advice that may be overhyped or just plain wrong? You know, what should we think about avoiding in conversion optimization?

Khalid:  I always talk about the best practices, the policy of best practices. Really, what works for one website might not work very well for your website. I recall working with a fairly large ecommerce company back in the late ’90s and they competed directly with Amazon. The methodology there was ‘let’s copy whatever Amazon does’. So as Amazon introduced a new design, basically our technical team would be working on that. While Amazon reported at that point in time a ten twelve percent conversion rate, that other company had less than a two percent conversion rate. So really, what works for one website does not always work for you. 

I’ll give another example. This was featured in the Internet Retailer website. One of the tests we did with one of our clients occurred during the checkout process. We added what we call an ‘assurance center’ where we said okay, what are the biggest fears, concerns, doubts that a visitor might have that would stop them from actually moving forward with the checkout process. For that one client, just merely adding a simple assurance center helped reduce abandonment rates on their checkout by about thirty-two percent. Since then, lots of consultants started saying well, an assurance center is the way to go. But the truth of the matter is we’ve had other clients where adding an assurance center actually caused reduction in conversion rates. So ultimately, yes, there are best practices but you have to see what works for your website and test it and then go from there. 

Anna:  What a great note to end on. I could not agree with you more , and that’s why when doing presentations or webinars, we’ve got to rely on those practices because we’re spreading the word, but always tell people to take them with a grain of salt. Just test and experiment on your own site because there are definitely no hard and fast rules when it comes to optimizing conversions.

Khalid:  Definitely. Definitely.

Anna:  Well Khalid, Thank you so much for joining us. It has been a pleasure to talk to you finally, on the record, and I just want to tell our listeners to please check out Khalid’s site which is invesp.com. And we’ll all keep our eyes out for the book you’ve got coming out with O’Reilly on conversion optimization. That sounds really exciting.

Khalid:  Thank you, Thank you. Pleasure all the time.