Categories
Online Marketing

SFBW19 – The Importance of Proper Measurement Within Digital Marketing for Crypto Companies

My name is Jamie McCormick, a managing director of Bitcoin marketing team. It’s going to flick through this very quickly here as well, Who I am been working in crypto since 2014 um. In terms of how I got here into the crypto space, so I classically trained in marketing – remember the Marketing Institute. I worked for 12 years in the article games industry prior to getting into crypto, specifically in virtual currencies and then five years in blockchain as well.

I ended up designing primarily because of advertising fraud. Three tracking and accreditation systems did one and article games have one in Facebook and then a generic platform as well, and we’ve integrated the tracking and accreditation methodology that we use in to over 10 crypto projects as well. I managed about 12 million advertising across this, so while you’re here um just to run through kind of the topics that are going to be talking about.

So we’re going to have a look at the smart approach in terms of a marketing plan. The importance of measuring how modern, tracking and Accreditation works some of the different challenges that are there to measuring the pitfalls. If you’re relying exclusively on Google Analytics the benefits of effective tracking a few examples, what happens when tracking goes wrong and then we’re going to have a look at a real-life example of a conversion funnel and specifically, how measurement improve that so in terms of being smart? For any marketers in the room, probably familiar with this, it’s an acronym for a specific, measurable, achievable, realistic and timely within digital marketing.

Taking a smart approach, that’s a cryptic company really measure all of the different activities that you’re doing to see what’s working. This is good and bad, and it’s really important that you tailor this against your specific KPIs. Now every type of project is different. If you’re selling a hardware wallet, it might be ecommerce sales. If you’re doing an exchange, you might have to do onboarding process trading trading volume and then there’s various different types of KPIs that you have and what this lets you do then has really build specific tactics in relevant marketing blogs for your different audiences.

There you can measure them, so you can really find out what’s working driving traffic to your website, into your conversion, funnel specifically through your conversion funnel and on to sales, you want to have it so that it’s achievable with your resources, skills and budgets. It’s realistic in terms of your expectations against your company KPIs and that you can find your audience in the right place at the right time with the right message now, when you combine correctly implemented tracking and accreditation, you can see things for their true impact, so good About you know not every advertising campaign is going to be brilliant.

Not every advertising campaign is going to be poor. You know often you’re going to have situations where you have some good, some bad. You might have say 100 publishers sending you traffic from a particular advertising network, but there’s only actually three or four of them that are really generating most of the sales that you’re getting in there as well. So once you can see, what’s good, you can focus efforts.

There once you can see, what’s poor, you can cut it off and redistribute that budget around then as well. So in terms of measuring you know anyone who’s using digital marketing. It is one of the most measurable activities that a company can do and you need to have your tracking in place. If you don’t have tracking in place, you can never really measure anything other than at kind of a generic overall level as well.

So when you’re trying to do digital marketing, tracking and accreditation, you need to have your platform configured. So when traffic is hitting your website you’re adding URL parameters, so people may be familiar with UTM tags like feed into Google Analytics. You can add a lot of costs and values onto these as well. These need to be fed into the link prior to being sent to your website or to an app store as well.

Once you have that in place, then you need to be able to take that data feed it into your own user registration database and then pull those reports into a marketing report which is limp, lined up with your KPIs and then. Finally, once you have it, you can act on that report and we’ll have a look at this. So knowing what’s good is great because you can focus budget there, knowing what’s bad lets you avoid wasted, spend so, if you’re running an advertising campaign for $ 10,000, you should know after about a thousand or 2,000 told us, whether it’s actually going to work or not, And if it’s not working cut it and move it move the budget elsewhere.

Also having the data as well has a lot of applications beyond it. You can find problems with your conversion funnel you can find some strange behavior, that’s going on anomalies. You know in article games, we found all sorts of products. A product design, you know, might be specific points within a game that were deep into the game that were actually broken and in a conversion funnel in the example will show later on.

There was two points that we’re losing over: 60 % of the user is going through. So when you’re dropping a lot of money on campaigns and they’re, not really going the whole way through, that’s going to cost you a lot of money as well. Now, once you can see what is working and you can fix it, you can focus and improve over time. So in terms of modern tracking and accreditation, you know, overall, your traffic is going to be mixed up from a couple of different sources.

So you have your organic traffic, which is going to give you a baseline, that’s coming in, so this is coming in from your SEO. This is coming in from any media that you’re doing in terms of peor. You’ve also, then, got traffic, that’s being driven from your communications. So this is your blogs, your social media posts and then you’ve got your advertising traffic than as well. Now I know our hosts ears were podcasts as well.

So, equally, you can use promo codes where you don’t have a link to measure and link it into this sort of a methodology as well. Now. The key thing to remember here is, if you don’t tag communications or advertising activities, you’re, never going to be able to measure them after the fact, it’s very, very, very hard after somebody has registered to then try and find out where they came from. So if you add the information when you’re sending it whether it’s through Google AdWords campaign, whether it’s through a ad campaign with someone like coin Zillow or coin traffic, podcast social media posts, you can do that in terms of P or obviously pure they strip out.

Quite a lot of the links, but you can use timestamp information to pin a cross-reference when traffic came in and link that back generally to particular types of activities, and especially when it comes to communication. Measuring, like we’ve integrated this with several of our clients of community management teams. So when they take kind of a unified approach to measuring activities, they can say: okay, well, we’re working across Facebook, we’re working across Twitter, we’re working across medium, we’re working across YouTube and what they can find then is well.

Twitter is actually getting us five times. The amount of traffic Facebook is getting us a lot, less mediums, getting us a steady amount and they can use that then to kind of repurpose their time in their effort. So they spend less time on the blogs that are giving them less results and they focus more on the ones that are so in terms of some of the challenges that you have to measurement as well. Now I’ve managed a number of marketing teams prior to getting into crypto, and human error is one of the big ones.

Now going back to my games career, we had about eight different people on our team. None of them were trained to the same standard weed people in from South Africa. We people from Ireland from England, Italy, Poland, Russia, Germany, everyone was digitally marketing qualified, but no one had the same training. So we really had to give that standard training to the entire team so that they could do it as well.

I blocked plugins really caused a big issue, so you’ve got the likes of brave where you’ve got default are blocking coming in. So in this particular event for SF BW, we were running campaigns with brave and because of adblock and because they were reliance on Google Analytics. We found it very very, very hard to actually measure the results. We can see the traffic coming and it was very hard to cross-reference that, because of this in terms of script, locking plugins it’s the same approach, you’ve got cookie blocking as well so say.

For instance, with here we had to you know you had a an e-commerce tag or a user ID that was being set by Google Analytics, and this has been shared into the Eventbrite and that wasn’t being passed at some point because of cookie blocking. That was going on now. Also in many cases, you’ve got advertising blogs that won’t or they can’t provide optimization data. So if you have a network that you’re talking to – and they won’t either give you a publisher ID or a website domain for the traffic, that’s coming in, don’t work with them.

You’ve got no way to optimize it and also, if they’re doing that they probably have something that they’re trying to hide as well. You’ve also, then got under reporting if you’re relying exclusively on third-party analytics platforms. So, generally, in non crypto space, you might have a 10 20 % discrepancy between analytics data and your own database encrypter that can be up to 40 %. I’ve see people more privacy, conscious people using more adblock and and so on.

Now that feeds into relying exclusively on Google Analytics, so we’ve worked with lots of clients when it came to analytics often they’re set up in terms of there’s a tracking script set up on the page and that’s recording, but often they’re, very often misconfigured. You don’t have event tracking setup, you don’t have conversion goals set up and it just makes it very difficult for you to read them as well.

As we said, there’s underreporting up to 40 %. You’ve also got a very separate at separation of analytics from your underlying databases. So, if you’re running an exchange, you know your exchange database is all of your information there. You might not necessarily be feeding it straight in so analytics is always going to be sitting on top now, if your analytics is showing less, then your real database, that’s normal.

If it’s showing more something’s wrong and again it’s misconfigured. If you don’t have on sorry, if you have on tag traffic as well, this can, combined with your organic traffic, can really skew your stats and really, you know the whole points of tracking is really building two different things. You know you’ve got the John Wanamaker dilemma, which is half the money I spent on. Advertising is wasted, I don’t know which half by doing tagging you can see exactly what it is and then you’ve also got the period or principle, which is the 80/20 rule as well.

Now, when you get tracking right, it lets you see at a very holistic level. At all of your marketing communication activities side by side in one report, so this can be your own team, stuff, organic traffic so coming in from pure stuff coming in from third party agencies and everything else, and you can measure them and I’d like for like basis. The approach that we’ve always taken has meant that you’ve got one report that you look at from your own data and then you can reconcile that against 30 or 40 different reports from platforms and see if there’s an AMA lease there as well.

So if there say they’re measuring lots of leads and you’re, not that’s one thing: if they’re measuring lots of traffic and you’re, not that’s another thing there as well. Now, once you have the tracking in place property, you can identify positive traffic sources and really focus your money and your time and your effort on these as well. In some cases, you’ve got it’s very hard to get volume and quality.

At the same time, and often what you’ll have is you’ve got an overall blog and within that you’ve got a couple of traffic sources, but generally over and over and over, are bringing in safe bets. So once you can identify a safe bet and traffic source, you can focus your budget there and maximize that out. You can identify negative traffic source, so these are ones that are bringing you in nothing they might be bringing in spent, they might be bringing in users or registrations, but they might not be going the whole way through your funnel and you can filter them out.

Now. Often, what we’ll see is you have a blog where you have some good traffic sources and then a lot of bad ones. So in one case we’re running a campaign with the clients they had a base. I think 60 different publishers that were sending traffic into the campaign two of them those publishers were generating 80 % of the traffic and spent on that. So when we tried to you know the other fifty-eight we’re bringing in lower volumes and they’re bringing in registrations, but when we cut those two specific traffic sources out, the whole campaign improved spend dropped, but the conversion rates all went up, click-through rates went off, and revenues Came up and the oral I came through from that as well.

So really you know it boils down to saving money. You know you want to save money on marketing spend also when you map out your conversion funnel and again, we’ll have a look at an example of one of these. In a moment, you can really identify bottlenecks in your process, so we’ve worked with lots of teams where the developers made the conversion funnel that they thought was okay, but in reality it wasn’t the most intuitive for end users.

So, by mapping this out mapping at every single step, looking at ratios between them and finding out what was working and what wasn’t we were able to identify, drop out points, some small, some big. So once you also have the product as well and you’ve identified that there’s a problem, you can identify barriers to product usage. So in a lot of cases you have to fully go through a funnel before you can use the platform and if you’ve a barrier there.

That’s stopping people getting into it. You need to be able to improve that there as well now development teams, all the ones that I’ve worked with know of data. You know the only thing that will get over an attitude in terms of oh, I built this and it’s great. What’s the problem with it? Yes, you did build it, but there’s a problem there. The data is telling me the problem. It’s not me. It’s there’s! Actually, a problem there, that’s there as well now, once you have it and you’ve got the recording in place, it’s very easy to quantify the changes.

You know you put a fix in on a particular day. You’ve got the stats all the way up to the previous day and then, from the day on you can see the improvement if it goes up great, it’s going to make your life a lot easier, the next time you’re looking for problems if it goes down or There’s no change, you can roll back and look at something else, and really what you want to do is by building this approach and really mapping every single traffic source against your KPIs.

You can look at the metrics there. You can find the certain traffic sources have higher metrics, so we’d often see you know, say with an affiliate. You might have an affiliate that has very low volume, but they’ve got a high click-through rate and a high product usage, and that might be coming from a texts post or some information about it there as well, whereas doing banners. You know you’ve got a lot more volume, you got a lot more traffic, but the quality is very poor.

Now, equally, when we look at say some of the crypto ad networks that are there, the standard banner sizes that are there have very victory rates, but some of the custom html5 placements that they have, which are much more prominent costs a little bit more. But the kicks your rates are five six seven times better and when we kind of measure that I you’re paying on a CPM that makes a big difference.

So what happens when TAC tracking goes wrong? Firstly, you can fail to identify systemic problems within your product design. It’s buried there, people don’t look at reports or databases in a lot of cases there and they just continue working away on something there as well. You can wait finite marketing budget on activities that are generating no or poor results, and they also bring down your overall numbers.

You can waste a lot of team effort, say if you’re have a communications team, that’s working across six or seven different blogs and you’re, giving equal effort to them all, but there’s only three of them that are generating what you actually want. You know you can save a lot of time and stress and hassle for them just focusing most of their time on the three that are working in terms of skewing stats.

If you have a lot of traffic coming in or you’ve got a load of poor results. This can really skew your stats and you can’t kind of change that then, and you can really focus on traffic sources, which you know you have kind of situations where you’ve got middling ones. You might have users that you’ve got decent traffic coming from a particular publisher or a keyword, for instance, but they’re only getting halfway through the conversion funnel and repeatedly there as well.

So while it looks good if you’re looking at the first KPI, which is an initial registration, but you look at say, KPI for they’re, not getting past that point there as well, and sometimes you might focus a lot of effort in time and money into those campaigns. Because they look good from their initial KPI, but they’re not actually going the full way. Also, what can happen is, if you take, you know, a broad sweep approach.

You may have a marketing blog. Like the example I gave earlier on, where you have one or two bad parties in there as well, and you cut the whole blog because you’re getting rid of those one or two so being able to pick out and pick and choose and really see, every single Traffic source compare them like for like and then surgically remove the one or two that are causing problems. That makes a big big, big difference there as well.

So it’s kind of seen the wood from the trees. Also, you know: we’ve had situations where you’ve got a lot of ad fraud. This could be amplified because, if you’re focusing on the right the wrong areas, you can have unauthorized incentivized traffic. You know faucets are kind of the bane of the crypto industry when it comes to advertising they’ve got and by the elements in terms of the onboarding, but generally you’ve got a lot of odd for that’s, going on and they’re from humans now coming on.

To kind of a real case example for a conversion funnel, so we have a client sees in the crypto finance space. They’ve got a complex seventh set formal and you had to get through this before you can use it so, starting off we’ve the traffic this had to hit one of their landing pages, and then we qualified the traffic with the information on this sent as well. They then had quite a long ball t-step registration process.

They they’re not to activate their email accounts. They had to set up two-factor authentication yet to enter in some personal data. They had to go through the KYC process, which everybody loves, and then they had to accept the program terms and conditions, and then, after that point they could get into the platform and use it now. This is similar to an exchange, but there’s a lot of platforms and crypto space that have something similar to this as well.

Now we implemented tracking with these guys a year ago and we’ve put tens and tens and tens of thousands of leads through this system and this approach in this methodology. Since we’ve been doing this as well and we were able to map out every single one of those steps, look at the ratios between them all as well, and then work with their teams implement the improvements. So in terms of what we did, we had the landing page, so we redesigned their landing page.

They were sending traffic to the homepage. We built called a sock or a dead-end road landing pages, and then we got a nice design that we were happy with. We did a B testing using Google optimize to kind of see which ones were best performing, and then we read at the traffic and focus the traffic into the pages that were working best with the registration funnel they had about 12 or 13 different input fields that People have to fill information in.

We really just search for those like what do you actually need, and that was about 6, so they could still put that information in monster in the platform, but during the registration it was that we also reorganize. This was a bit more intuitive for users as well in terms of the email activation. So one of the problems that they were having was a lot of their activation. Emails were getting hit by spam filters, so we got DKIM authentication and this improved email deliverability.

We also rearranged the subject line, so it was cleared. You need to click this to continue and that brought up between step 1 registration and step to email activation numbers by 6 percent. In terms of the next step of the personal data and the KYC, we work with them to rearrange the process and also they had to kind of re engineer some of the elements with their kyc provider. But once we did this, you know they were losing in 52, sorry 48 percent of people weren’t peak, passing kyc.

They were just getting to that step, looking at it and going no. This is too complex and at the end of doing this as well, that went up to 75 percent, so we improved the numbers by 35 percent just by reengineering that also what they had a two-factor authentication step. This point here was causing a lot of problems. They were losing over 30 percent of people. So if you do the maths here, you know you’ve retired some people coming in.

They were losing six percent at Step, two at Step: four, they were losing 30 odd percents and then it stepped five or six they’re losing 30 percent there as well. You know the funnel there was losing because of its design over half of the people that were going through. Now we had to work with a marketing team, their product management team, the development team. But overall we did it led to an improvement of 40 percent of customers gone through the process and from the same spend and the same activities.

We brought their sales up by 20 % as well. We also then caught the blogs they were using by 50 %. So that in weed less overhead time effort whole shabang there as well. So that’s the end of the presentation. We have a booth inside in the hall we’re offering an heiress free consultation and a chance to win a nice bottle of Irish teens whiskey here as well. So if you want, if you’re interested in having a chat about your project and tracking an accreditation we’re more than happy to have a child with you next door, so thank you very much.

Everyone for your time and attention appreciate your listening. Sure hey. Does anyone have any questions, so I’m black and blinded here, so I can’t see no fair enough. The same approach you can map in serve any sort of yeah. It does so we learned this in articlegames. You know, and article games had a very complex. You know if people get to register, they had to download six or seven gigs, which you couldn’t track.

Then they’d have to login, go through character. Creation play the game, go through all the products there as well. So the same approach works for the onboarding process, but then equally it works for applications as well end to end so cool okay. Thank you very much, oh okay. So so the question was what what programs, what programs you used to track user behavior analytics yeah? So the approach that we’re doing can integrate in with any of those visualization tools or analytics it’s more methodology that feeds it straight into your database.

So there’s a bit of coding: that’s required at the start, to get it integrated in your site and then usually what we do is we’d have sometimes use a Mongo database, some people using Splunk some people using SQL once the data is in there. We structure the reports and that can feed into any sort of visualization tool that you have so the approach that we’re taking there’s no like analytics pixels anything.

It’s just there’s a methodology that once you get it in it’s in there and then you can then use whatever tools. You want to use whether it’s a CSV report or an SQL makes panel whatever. It is actually visualize that data okay, we’re going to have to leave it there. Thank you so much Jamie


 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.