In this, our fourth episode of Positive Chats, we interview Gemma MacNaught who is a UX (User Experience) and Conversion Rate Optimisation Strategist. With Gemma, we discuss the art (and science) behind CRO, what this actually means, why it is necessary to test the user experience of your website to determine how users interact, and how this can be applied to achieve your online goals.
Embed not working for you? View on YouTube!
If you want to find out as soon as we publish the next one, why not…
This MailChimp shortcode is now deprecated. Please insert the new shortcode to display this form.
Jon: Hello and welcome to the latest edition of Positive Chats, today we will be talking about conversion rate optimisation with an expert on the subject Gemma MacNaught.
Jon: So let’s get down to brass tacks. What is a conversion rate and why do I need to optimise it?
Gemma: Okay, a conversion rate is the number of times your goal was completed over the total number of visitors, times by 100 and gives you a percentage. To optimise that there are various ways and I’ll start off with the definition. When I first got into the industry there wasn’t even a lot of information on conversion rate optimisation if you Google it now, you will get swamped by it. So the definition as Google tells it in, it breaks out to creating an experience that makes it more likely for users to complete their goal and what’s interesting about them using experience is there still seems to be two schools of thought in CRO, Conversion Rate Optimisation, and that is that the first experience is the controlled environment you create, so we want to setup a test where we have changes that we want to try against an original by setting up a testing environment you are controlling the factors making sure they are both subjected to the same things like traffic.
The second definition or school of thought is experience as creating a digital space where users are more comfortable, more encouraged so your website as a whole product, so it’s much more user centred much more resource centred. The end result is to test them to make sure that you are doing what you want to do and your user to do and what you want them to do but really it is much more focused on achieving user goals and that therefore achieve business goals.
Jon: Sell it to me in one sentence.
Gemma: Conversion rate optimisation, I think the best way to put that across is the ultimate way to achieve your online goals.
Jon: Excellent, why is it important and what are the benefits?
Gemma: Your website is hub for all your marketing activities so years ago and even in some cases still today, people will still have physical stores, but many people don’t so it’s ridiculous space and it’s a representation of your brand. We have about 40% of all population online right now. That’s a 700% increase since 2010, and it only increasing especially with more mobile uptake in various country overshadowing desktops even. So more, and more people are going to be online, they are going to be trying to access your brand so you need to make sure that it is performing properly. Google is, you want to keep them happy, as you know is all for optimisation with them, Panda and Hummingbird, that has been going through the Internet stripping down pages that are low quality content or not delivering on promises. With optimisation, that’s exactly what you are doing so you are keeping Google happy so it’s going to help your SEO rankings as well.
They have also just started recruiting for manual researchers so it’s not even just going to be algorithm anymore there is going to be actual people going through the websites, making sure that they are up to scratch. Further benefits there are, so your SEO spend, your marketing spent is more valuable, you are converting more visitors into customers so your growing market share and for every customer you get, your competitor gets one less, so really all round it is a very good business to be optimising.
Jon: I have got to ask, is this just a fad?
Gemma: No optimising the market has been around since before slice bread and that’s a fact, the slice bread which came out after, scientific advertising which is book by Claude Hopkins who released that in 1923, he had been practicing optimisation in his market for many years before that. This was print advertising so he realised that there were various things you can do to measure your efforts, various ways to get customer more involved. The first one was he made his copywriters research the target audience and got to know them really, really well and then put the ads in with different headlines with coupon codes, they had a slight difference so he could track which ones did best, and the end result was better use of resources for his clients they weren’t wasting so much money on their advertising and for very little return they were putting in less but generating more business from it.
Jon: Is it any different to UX?
Gemma: Yes, and no. If you go by the first definition so CRO is just testing then yes, it’s different completely. If you go by the second definition that testing is a tool, and that there are various other methods for optimisation and research then it’s very, very similar. I was at a UX conference where I got treated as a bit of a leper, when they heard that I was in the business of conversion rate optimisation not UX but the more I got speaking to them, the more they realised that I was doing exactly almost the same work that they were, it was just that at the end of it, I did a test, I split tested my changes to see what the increase was, to see the impact so really a lot of the methods used are exactly the same but I would say that I have come across some UX who actually do split test which was great to hear, I don’t know why they wouldn’t but if they, that side of the industry doesn’t catch on more you could probably call it, ‘Conversion Rate Optimisation’, like ‘UX on crack’, something like that because you are really getting the results, you know you get a figure for your efforts, you know what increase you are having.
Jon: I think, I am going to use that actually in our marketing, ‘UX on Crack’.
Jon: I think that’s what we should go with.
Gemma: A lot of the time, because conversion rate optimisation is still a new term to many people I prefer to refer to it as user experience optimisation just because of the similarities and people understand almost straight away what that is so yes, it’s user experience optimisation that I am in the business of you can get so far with best practices for your testing but if you really want to improve the quality of your website and really make push it, sure that’s it’s working as hard as possible you need to do your research and you need to get to know the users.
Jon: That’s just far more advanced and more specific to each client.
Jon: Okay, so onto the next question are there any basic rules of testing?
Gemma: You need to start by defining your goal, every website, every marketing campaign has a goal and it’s measurable, once you have that, you do your research work out which areas of the website it is that require the most attention. Start there first of all get yourself some testing software that can be Google Analytics or there is Optimisely or Visual Website Optimiser, there are lots coming out now but if you start with one of those basic three and work from there, if anyone comes out, have a look at it but those are your sure-fire options. They are very reliable. The testing software that you use, splits your traffic out, so if you have one page that’s your original, you have a second page that has all your changes on it, it splits the traffic between them and tells you which one is performing the best.
Basic rules, you will probably see like a peak, when you first start running a test but you don’t stop it there, you need to leave it to run, even if you are being shouted at by a manager saying, “Oh it’s doing great, just take it off,” don’t do it, you need to wait until at least a hundred conversion I would say, per variation before you can decided that this one is definitely there, the testing software as well will calculate when the result is statistically significant. The reason that’s so important is because if you stop a test too early it could have been a fluke, it could just be a bump so when you stop and implement those new changes, it wouldn’t hold the result so make sure it reaches statistical significance, then you know that when you switch to your new changes they are definitely performing for you. Then the time it takes a test to run depends a lot on your traffic and your distance conversion rate, there is calculators out there that will tell you how long you can expect the test to run.
Jon: So I have heard of such a thing as over optimisation is that true?
Gemma: I have heard that too, and a lot of people would say there is but from my perspective there can’t be such a thing, ‘Over Optimisation’ because the results we see coming out from timely changes, like on buttons, one word can make all the difference how do you know whether you are not pushing it, just that you are not, you are sitting there holding back from making that tiny change because you say, “Oh we are done, we are done optimising, it’s performing great, we are leaving it there.” You can make that one change and get great result. Also when it comes to a full site there is always going to be new products, new campaigns, new services, seasonality. There is always going to be new things, new things to try, and new things to test, so you know I don’t believe there is such a thing as over optimisation.
Jon: So why should you test why not just redesign?
Gemma: There are a number of issues with redesigns, and that is that you are just implementing a brand new website, you may think that this one is so outdated now we need a completely new one, let’s just do it but the issue with that is that your old one could have been performing quite well and you have just implemented a whole new one and completely changed everything how are your users going to respond to that? So one way to look at it is you want to buy a new car, the one you already have is a better dated, maybe it’s not working as well so you have decided to upgrade, you go to the car dealership, you are going to take that car for a test drive, you are going to make sure that it is better than the one you have already got.
You don’t want one that’s worse, and then after that you are going to maintain it, you are going to make that it keeps its performance everything like that so you have this respect for cars and unless you are a race car driver it’s not going to get you as much money as optimising on your website so we should put the respect there also. So definitely redesign there is a case for it, if the website is really badly built in the back end, if it’s seriously broken, if it really doesn’t represent your brand, but even then you can still do all the research, you can still find out what’s important to your users, and then use that, apply it to the new website rather than just starting from scratch ignoring everything that gone before which a lot of people are still sadly doing but with testing, you really are limiting the risk, like it is a test, so you are putting in this environment that makes it safe to take risk and to make better changes which is if you really want to redesign your website you can get to that point anyway through optimisation.
Jon: So I have got to ask, what is the most surprising change that leads to the most impressive result?
Gemma: For me, it’s hard to pinpoint just one but my favourites are the small changes, the headline, the colour of a button, adding a testimonial, the results you can get from things like that are phenomenal, really surprising and those are great tests to do when you are just new to optimisation because it really show you what can be achieved. So there are several case studies out there for increases like that that are really, really tiny and we haven’t had to redesign a whole page, it’s not a big build.
We have one the Render Positive website right now and that was the social workshop, we had to work with an existing popup normally always a bad thing, and use that to generate newsletter saying “I support them.” The content state exactly the same, there was nothing new added it was just re-authored, we emphasised stuff in the headline that wasn’t before and we reordered the buttons that was already there, those small changes lead to over the 500% increase in our newsletter signups, so those for me are the most surprising, because it’s minimal effort and they were shocked themselves that something so small could yield such a great result.
Jon: So every client wants to know are there any universal quick wins that they can implement?
Gemma: There are best practices that you implement and if someone in your industry has a case study and they have got a great increase on changing a certain part site or adding a new tool you can try that but I would say, “No” there are no quick wins because you need to get to know what your audience wants just because comparative of yours is done, it does not mean it will work on your site, if that was the case every ecommerce site would look like Amazon, so it doesn’t work like that.
The best way to look at it, is that following a rule of website equals truth, so we have, ‘T’ and that’s trust, that is a really overlooked factor on a website, when your users come to you they have to trust you, they have, you have to look like you know what you are talking about and that includes presentation, layout, imagery, the content going onto ‘R’ so that’s deletable, you have to answer why the user should pick you and not go to your competitor, you have you, that’s usability make sure nothing is broken. I suppose if there is a quick win it’s if something is broken do fix it.
So a lot of people haven’t even been through their own website to make sure that the steps on the user journey are easy to complete and there is no massive red flags, so it takes someone like me coming in with a fresh pair of eyes to go through that and point that out to them but they could have gone to someone in their team for that small bit, where they just have a fresh pair of eyes looking over it for them and, ‘E’ is engaged so you want to give the users a really clear next step if they are not ready to convert and complete your goal, there and then, you need to find a way to hook them in, so that you can get them back to the site again whether that’s social or newsletter signup anything to get them interacting with the content.
Jon: You talked a lot about research what are the tools we should use to carry out CRO audit?
Gemma: The first place I go to is the Analytics, there are various paid versions out there but there is a free one, Google Analytics, and it is the most widely used but some people do prefer to go with a paid version. What I find is that means clients can find it difficult to interpret their own data. Normally they have an account manager with that company then have to tell them about their own data. So really I would advise people go with the free version, go with Google Analytics, it’s going to give you all the information that you need and once you start looking through that and get it all setup properly it’s really not that difficult to find what you and you can setup dashboard and everything like that.
So Google Analytics shows me what is going on on the website where users are landing where they are exiting, where on the customer journey or the funnel that they are dropping off, where are the top traffic sources, how they are performing, everything like that I can find out from there, so that’s my first point of call. Going on from there, it will be usability testing, and we can do this remotely which is really quick and really simple, usertesting.com I think it’s a four hour window and you are going to have results back and that’s really users from your target audience following a set of questions going through your site, and they will flag up any areas that they think need improvement or you will be able to see when they hit any difficulties. There is also …, being able to do that in-house, that is more difficult, that means that you have to setup a laptop, get the recording software everything like that.
The reason that I mentioned the kind of in-house version is because it transfers over to mobile and there is a lot of applications coming out now where yes, you can still do it remotely you can get usertesting.com to do a mobile test, but when it comes to mobile you really want to get out into the field, you don’t want someone sitting at a desk doing it, that’s not how most people are looking at their mobiles, so go out and have a look for mobile app testing software if you want to do that. What else would they normally look at? Page Analytics, we have iTracking, there is a really good service called, ‘iPhone’ and that has created an algorithm that will analyze a webpage and show you where a user’s eye is drawn to on that page. That can be a pricy option if you are a small business. The other option would be five second test, it’s where you upload and above the full version of your webpage, ask a few questions, users get to see it for five seconds then it comes back with answers.
The best question to use are stuff like what is the brand, who are they, who are they for, what was the next step on the page, was there any way to contact them, really you want to make sure that the first impression is good like just with anything really so Eyequant and the five second test will give you that information, and going on from there survey is an amazing way to gather feedback from your user, you can either do that onsite with services like Qualaroo or you can it through email with stuff like Survey Monkey and one of the good things is that everything I have just talked about so the surveys the page analytics, oh, also Crazyegg for tracking clicks, have been kind of rolled into one now by some platforms so we have stuff like usability tools, and Hotjar and these are great because you have got your URL you put that in there and then you have all these option of how to analyse so those are really really useful and it means you don’t have to keep logging into different services, they are all in one place.
Jon: So the big question is CRO necessary for every website
Jon: We spoke a lot about the ecommerce sites, but should smaller, lower traffic sites use CRO, would that benefit them?
Gemma: Definitely, if anything it’s even more important for them because they could have some pretty major competitors so they have to perform; they have to put themselves up against big brands. There are ways to test on low traffic websites, there is things like focusing on micro-conversion so click-through rather than end goal, we can also test really big changes, so going all out, testing site so that we use the maximum amount of traffic to reduce the time a test would take. We also have, we can split the traffic differently so when you would normally run a usual EB test, it’s 50% traffic per variation, but we can maybe make that an 80/20 split to the new variation, that’s also really useful because it just shows you automatically if the page is going to do any good or the change is going to do any good. You don’t want to put a change live that is going to negatively impact your business and that is a really quick way even if it doesn’t reach full significance at least you are trying to do your research and also the research there is nothing stopping them from going and doing their homework finding out what it is that the users want because that is going to benefit them even if there isn’t enough traffic to test, or they don’t want to make a time commitment to it, because it may take a while for it to complete, if you did the research and let that influence your changes and your updates, you are going to be going in the right direction.
Jon: Everyone has a Smartphone these days; do companies need a separate strategy for mobile?
Gemma: That has two sides to it, so there is a shift now towards responsive design which essentially means, it displays differently but the content across should be the same, or there is still people who have mobile websites and desktop websites. If they have mobile websites and desktop websites, the strategy is going to be completely different, if it’s responsive changes are transferrable but obviously we are working on a time canvas now so it has to be finger friendly, you have to consider that the person is not sitting at a desk, the general setup of a computer is desk, chair, screen, keyboard, everything like that. On a mobile, you have so many distractions around you, it can be very time sensitive, you might just be on the train so all these considerations do create a new strategy. While content should be accessible on both mobile and desktop and tablet, its how we access it, how we prioritise it that needs to be different.
Jon: What are some of the most common CRO pitfalls that you see over, and over again?
Gemma: Oh, lots, the one I see all the time is where a brand has a lot to shout about they have won awards, they have been in the press, their customers have taken the time to write testimonials for them and given them five stars, but then you go onto their website and it’s nowhere to be found. You can’t find it, if you can find it, it might be hidden in the in the footer or somewhere and I can’t believe that still happens. Everytime I see it I am like, why are you doing …, because it can, seeing that straight away as soon as a user lands on the site is a massive trust element, it boost you straight away. Other things is using jargon so if you think that your website is industry specific it’s very targeted so you use a lot of jargon in there but that’s not necessarily key, people don’t want to have to work out what you are saying, they just want the layman speak really, they want to know, yes, this is exactly what I am getting without having to read between the lines.
Other things, irrelevant imagery, stock photos they are not always bad but most of the times they are, don’t use imagery that is completely irrelevant to the product of the service you are trying to sell, just because the Internet loves cats it doesn’t mean you should have one on your website because that can be utilised so much better even the psychology of faces and stuff like get a human on there, don’t put a cat unless your business is about cats obviously.
Jon: Playing devil’s advocate is there anytime when conversion rate is bad to use as a metric?
Gemma: Yes, there are times, I will try and explain it down as simply as I can just so that I don’t make too many enemies in the industry. Obviously I am a CRO advocate, but this is sometimes when I have an issue with the name, Conversion Rate Optimisation and would prefer to call it User Experience Optimisation because conversation rate as you said is a metric, it’s the, the goal of completion is over total visitors and that’s how you achieve that but if you try to measure that on a site, across a timeline, it’s going to give you varying results either way because you have so many factors influencing it and it’s going to go up and down so if your traffic goes up, your conversion rate could go down.
If your conversion rate goes down, that doesn’t necessarily mean that your sales have gone down which is a key business metric. So really when we are looking at conversion rates it is useful in a test, it is useful when it is in that controlled environment to make sure that a new version is definitely performing better than a previous version. That means that if you fire a whole load of traffic or campaigns at it, it’s still under the same conditions, both versions are getting the exact same visitors through them so we can measure the difference between the two, so if that is out and just, if you just made a change and then your conversion rate drops down, you may attribute it to that change but really it’s because a new campaign started that is irrelevant to that goal.
So your goal could have been purchases but you have just sent a load of traffic to your newsletter signup page, so that’s not the same goal, so that conversion rate goes down and traffic goes up, so you are, but if people are coming in and they are signing up for your newsletter, it may then go on and make a sale so the sales goes up but the conversion rate is not going to say that because of the influx of traffic. Does that make sense?
Jon: I think you may have made lot enemies in the conversion rate optimisation in this in history.
Gemma: Oh no. Let me try and redeem myself. So when it is useful to look at a conversion rate is when you are looking at users going from a page to a page or a micro conversion at that point in time. In the test we are measuring the difference between two conversion rates so you might have seen an influx of traffic, here is your original and here is your new version. That’s still up here so there is still a big difference there; you are still getting many more sales than what you are with this one. So when that test completes, you are still going to get more sales than that original but using conversion rate across time it fluctuates far too much, and you need to look at other metrics, ask yourself why your conversion rate has dropped or increased.
Another problem with just solely looking at the conversion rate metric is if you are looking at your traffic sources, so say email is doing really well it’s your top converter are you going to turn off all your other marketing spend, and just focus on email? You could do that but that’s not sustainable for the longevity of your website. It’s not a rational thing to do, so we can’t get too hung up on conversion rate. If you go into your Google Analytics or any other provider that you are using and you see my sites conversion rate is this, that has so many things contributing to it, and also depending on how you have got that setup that could be you purchases, your lead generation, your email signups, your video plays however you have got your goals setup within that data tracking software is contributing into that conversion rate, so if you look at that it’s not really telling you anything so yes, we have to increase our conversion rates but we have to measure and track it and understand what it is and why it is going up and down. We can’t just look at it and not look at any other data.
Before I forget there is one more thing that I’d like to go over. A lot of times, I get asked by clients what increase can I expect on this test, how much are you going to increase my conversion rate, and the answer is, I can’t answer that, nobody can, if anybody comes to you and says, “I will increase your conversion rate by 50%” they are lying because they don’t know. You don’t know that’s the whole point of testing, you want to work out is this change going to do this business good and by how much, so if you see something implemented by one of your competitors and they probably look at a case study saying that by turning their red button into a green button that got them 30% increase on their sales. If you try and do it that does not mean you are going to get that so it’s so difficult to predict, there is too many factors that need to be considered.
Jon: Thank you for answering my most annoying question.
Gemma: That’s okay.
Jon: If it’s so important, why isn’t everyone doing CRO?
Gemma: I think the main thing it comes down to is know-how and resources. A lot of the time companies will try to do it in-house and they will maybe do a few tests and they don’t get the wins they were expecting or they find it too difficult or their strategies are really disjointed it’s not very fluid, they are not learning from the testing and they don’t get great wins, Econsultancy did a survey in 2014, I think that was up at 80 to 90% of Gemmas said that conversion rate optimisation was at the top of their online marketing strategy again that will be bias because they will be targeting people who are very familiar with online marketing so I have come across a lot of businesses who don’t know the terminology, they do need it but they don’t quite understand what it involves and they may try and take it on in-house but they just don’t have the skills, and the experience to be able to see all the way through it, it a continuous process as I said, I don’t believe that there is a point of optimal optimisation that you can’t go any further and can’t push it anymore, a lot of companies do still see it that way where they might do one test, they will get a good result and they will go, “Oh great,” or they might do one test, they don’t get very good result and say, “This is a waste of time,” but with such high numbers in the industry so from that Econsultancy surveys saying that it is critical and more and more organisations are beginning to test their websites and research more about their users, everyone needs to jump on the bandwagon if you don’t then where does that leave you, especially if Google was now trolling through wanting to make sure you have the highest quality content and if you don’t they are not going to rank you as high so that then has an impact on your traffic so really everyone needs to be taking a serious look at their website and doing everything they can to make sure it’s working as hard as possible for them.
Jon: Well hopefully this video will convince more of you out there to do this. So I am all out of my questions have you got anything else you would like to add?
Gemma: As a final word, I think I will leave you guys with saying that we need to remember that it is all about our users, we need to get to know them because persuasion is being able to relate to them you need to answer the why and not just the what. We need to think of our websites as those digital spaces as a 24 hour representation of our brand, we want to make the experience for the people who come to visit us as seamless, as easy as possible and as persuasive so really, yes go out there, do your research, do your testing, measure everything and good luck.
Jon: Thank you very much for your time.
Get more great stuff delivered fresh to your inbox.