Segment: one API for analytics, tracking, and visitor segmentation

Hey guys it's Nathan today we're just gonna do like a little walkthrough of a really interesting tool it's a tool called Segment I'm sure some of you will have heard of it but if you haven't basically what it does is rather than needing to implement loads of different scripts onto your website and have like your developer perform like loads of different tracking let's say for example you want tracking of like when a user adds it to your cart and and let's say you're doing advertising on like Facebook on Google and Twitter on like LinkedIn and maybe you also want to send that data over to Google Analytics and you're looking at like a number of different integrations.

So Segment actually simplifies that by just let you only need to integrate their API and then from their dashboard you can actually sync over to what we call sorry to what they call like sources and and destinations which means that all of your data is sort of synced to all of your tools within within real time and it also means that you can push over like back to Google add to the exact checkout amount and things like that by putting the data in from stripe.

So it just allows you to make various different integrations between like all of the the website plugins that you use now it is like it's expensive is it's not a cheap tool to use but it is really really effective I had first used Segment about two years ago and it was actually like due to the price that we that we moved away from it because it just just wasn't workable for us on that particular business but like particularly for like a b2b company that has like quite high margins it's definitely like at all worth using.

So for your first like 1,000 visitors per month is actually free and then from here on it's like a hundred and twenty dollars per month per 10,000 visitors that you have and if you're running like ads campaigns and things like that the pricing does scale up pretty pretty quickly but nonetheless it's like a great tool to use the only reason we couldn't use it in that particular instance is because we were like a low ticket item but we've actually found that a Segment startup discount you can apply for relatively easily and we've been able to get I think we've got a whole year of segment access sorry twenty five thousand dollars in credits for free.

So you can actually go ahead and like apply to that I definitely suggest that you do and like you can see like all the different tools and integrations that they have there's just a ton of different sort of flows and ways that you can sync data across they support personalization which is quite interesting and like a ton of other things but as well like I said I've just just come back to it for the first time in my two years and if I can just.

So yeah we which ones we're tracking things via our website you okay here we go so I'm back in here so one tool that I like I was just playing around with now which I think like previously you had to like implement it via like coding yourself however they've now built this visual tagger so you can add in your tracking like really really easily and I think like this is just like just beautiful so socially I don't want to do that actually so I've just like implemented all these different events and we're just going to add like a few more.

So when someone initiates to check out on various different things like it took me like two minutes to to set this up and like we can go ahead and just add another event just to like show you how easy this was so it brings up like a interactive version of your website and the like I love this so much because it is actually telling us whether or not our script is installed and this UX is just like fantastic is actually really quite a simple thing to do however like how I've never thought of doing this in notifier it's like unbelievable.

So I'm definitely going to like implement something like this because we're also like a script based and SAS solution so just for someone to make sure that their script is actually sending data is so simple but it just works so well so currently what I've done is set up tracking on all of the Buy Now buttons so that we know when someone initiates a check out and now what we can do is just so yeah again I'm just gonna give you a quick demo of how like easy this is.

So let's do like when a user submits a form and we're just like is picked it up for do we want to track information user submits why not why don't we pick up the yeah so now okay this is just this is so beautiful and the funny thing is is it ties right into notifier because on my other company notifier we're adding in like personalization where you can also identify users or let's say like you're using a tool like segment you could send over to us it's this visitor currently on your website and any notifications that you displayed we would just add in like the name email or sorry we would add in the name to make it like personalized if at 59 I think.

I've got a hidden input field okay that's really nice so now we can go back to actually testing like you'll see they've got like a live it like it tracks in real-time what you've done so for example I can just go ahead and and fill out this this form I'll just pause the video quickly whilst to enter my card details cool so again went through and it marked as being successfully initiated I did actually make like a video on doing this like ages ago but I'm just going to give like a quick demo as well.

So for anyone who like hasn't done this before it is like so you can just track that as many things as you want on a page now in general just to give like some sort of context like this I think it was actually through it was a growth hacker who actually used to work at segments that I watched a video and learned this from so what you should do is keep like the most important data like at the top of the page and it was also to keep everything just like.

So that you were it's quite difficult to explain this if you haven't done this before but you sort of want all of your data to sort of stay in in order so it's like when you're sort of reviewing it like a server down the line so rather than saying like homepage button and then or saying that call-to-action button on all your different events you want to come up with like on a structure because otherwise when you come to review this further down the line let's say let me sync over to your like CRM or intercom or something all these different events they're going to appear in like alphabetical order.

So it's going to be like all over the place and if you have like 50 or 60 different events you're going to be like just spending ages trying to find them so instead if it'd be like button cool okay so yeah like they've given an example up here now so we would do like button homepage and if we're running like A/B tests we can do like button color purple like five hundred for example that's the shade of purple it depends how like intricate you want to go with it and button text watch demo.

So this is more for likely if you're running like a A/B test and you want to like check all the data of what's being clicked and the click rates and things like that so sorry actually I've structured that wrong it would be button color trying to so watch demo so you are assigning a value to like each of these events and it would be like button location like something like that so this is obviously just for one example and then when you you set up tracking of the next button like the button here you want to make sure that all the different properties that you're tracking that you're following this structure in how how you're tracking it for thing because otherwise like the data just.

Becomes pointless so you need to be really really strict in how you perform your structure I also like I've again done this a bit stupid but the most important data that you have should be up at the top of the page I saw it up at the top here so it's at the very top of your like analytics and again keeping it in like a structured format button color button text location rather than putting location button color and text so the data is all just gonna get and it's just not gonna be like a clean process for me to use it.

So this is just something that I learned and I still passed on to to my developers but on how to do it but now you can actually just do it yourself with essentially a drag-and-drop builder which makes things like super super easy another thing to learn with analytics that I've I think I've just wouldn't have learned over the years is don't don't track everything work out like what what are the most important things to your business so again if we're working in software as a service we should be working off like acquisition and activation revenue retention and referrals in my opinions that the best best framework to go with.

So even just tracking like one metric within each of these within each of these sort of stages of your funnel so for like acquisition what are you counting is your so it's activation to acquisition so what you counting is your is your acquisition metric so like I I wouldn't count it it's like all website visitors in my opinion that's as pretty is it's not a good way to to track things like maybe I'll get up like a Miro board if we're going to gonna go through all of this stuff.

So now we just so we're going to have like all the different stages of the funnel right so the first thing is acquisition so users coming onto your website but we don't want to track like every single website visitor so we want to do like sort of segments so you want to segment out your returning visitors and as well I would I would also segment out anyone who doesn't spend doesn't want to work for me so anyone who doesn't spend at least like 10 seconds on your website or like 15 seconds I personally would would filter them out because it means is that they're not the type of people that you're trying to target.

So you don't really want them within your like analytics if you're going to be making decisions around like this data that so removing people you bounce within 15 seconds or within like 10 seconds whatever you choose to do I think it's quite a good thing to do of course thank if you're focusing on optimizing your bounce rate you're not going to do that the optimizing of bounce rate isn't really a good metric to this we'll start off with like further down the line you can run like A/B tests where you you could try and increase the amount of time people spend on your page but again this is we're performing like full funnel analysis here.

So we don't want to add like every one into into our metrics we don't want to be performing that data analysis on people who aren't relevant to us because then you're going to be making decisions on on how to run your business based on this sort of people that we aren't interested in optimizing the business for and so I hope that makes sense so that there may be some other things like in your industry as well of like returning visitors I would be for like segment them or like if you have a way of tracking if they're a user or not you don't you want to make separate that because they're already further down the funnel.

So make sure that you have like a clean template for your your analytics so like acquisition stage I would actually say like quite an easy simple one to do it could be like visits pricing so if you have like a separate page for visiting your pricing page and that is a really good indicator that someone is I interested in what you're doing if they're visiting pricing page or if you have the pricing page on the home page and you can't use that as an indicator I'll do something like visit signup slash checkout page.

So whatever comes first in your funnel if you take someone to a signup page to make an account then I would use that as an activation metric if you use the if you redirect straight to a checkout page which we do if you're like ecommerce then that is a pretty good good metric to track so again with in commerce and it just changed it to blue so the funnel makes a bit more sense nice of blue so these are things that we don't want to track we're not interested in them these are some ideas for things that you may want to track in your acquisition stage based on on your business and the structure of your website again it can vary there are other things.

You could track as well like how many people spend at least 30 seconds on your website how many people spend a minute on your website could be treated there's a a good like acquisition metric and yeah but then they're like we can get so much more complex into this in terms of like where's the traffic coming from but I would implement all that like further down the line once you've got comfortable with your analytics and understand understanding the data you can get much more like granular with the with the data that you're feeding yourself but to start with try and make it like really easy and digestible and because really you just don't want to overwhelm yourself of data at the start.

So again like this funnel is going to like vary based on based on your business like activation could be treated if you're like a SAS company it could be treated as a free trial it could be yeah it could be like a ton of different things and but I'm most of my the people who watch my videos have access founders are involved in like the SAS business rather than my e-commerce so let's basis on like a software-as-a-service framework [Music] okay cool so this is just generally like the the beginning metrics I would start tracking.

So so for activation I mean technically we could also put it under a pricing section it depends what we want to track and like we have quite a small websites there's more like limited data I could even change the order of this round because we're not offering a free trial we're actually charging up front okay okay yeah so this is the metric that I'll track in this scenario how many conversions let's say let's just do that under activation revenue we're because a SAS subscription like the biggest revenue is actually going to come from like renewals therefore revenue there are three different metrics that we could go with it's going to run the same metrics all right.

So we can do customer lifetime value it's going to be a pretty good one and so I'm not suggesting that you track all all of these these are just like some ideas because it's going to vary depending on your business another metric that you could actually throw in there alongside that is percentage of upgrades so renewals with this has changed at to churn but so how the churn is how many users are canceling every month for lifetime value is what's the average amount of money that you make per customer revenue is going to be like this and this.

So percentage of upgrades is going to be what percentage of customers upgrade onto let's put that on and like retention again is going to be really like based on your business like we could run retention based off usage is going to be a really good indicator of how engaged someone is within within your application so this is just our way of trying to like reduce reduce subscription we could also put lifetime value under retention as well essentially of upgrades in fact I'm so we can do like the lifetime value and then this I guess is more like an action metric.

So how many so by tracking these two metrics we can actually we can cross-reference between lifetime value and also reference based off how many people cancelled like if they had them low usage the previous month how many times would they come and visit us so on and so forth luckily for us we have an API so we could see exactly how how many times someone used our application within a month and you could like sort of piece this data together to try and understand where users are if a user gets below like a certain amount of usage or doesn't visit you in 30 days you probably want to put them into like they're at risk of cancelling mark and then add them into like your.

Email automation sequence or something to try and sort of resurrect them as a as a customer again like this another point I've missed is so it's getting late here and I'm getting tired so resurrection is how what percentage of your users who cancel how many are you able to bring back so entering like a resurrection flow is always a pretty good idea and then just finally isn't it as a term for how many referrals just one customer bring so if you can get your business where like every for example this is a very famous Dropbox growth hack this is how they grew.

So quickly I think I've read at one point they had it at a rate that one use it would bring on and enough of two users on average it could have even been free I don't quote me on that I'll have to to double check it but it was it was something in that region so every time they brought on a new user they were pretty much like doubling their their user base and they just had like this very easy viral effect so if you're able to implement something like that this is going to be like a really powerful way for you to grow it.

So just to might go off this again once we're tracking all of this data your goal as a growth hacker should be how can i optimize this funnel so i suggest working in like sprint which is a software development way of working but you would usually when like seven would be quite a short sprint but in growth marketing is quite easy to do that she would run like a seven-day sprint working on each different metric that you have and working on how can you improve it.

So whether you're not on your sub by yourself or you have a team on day day one i would actually that constantly referring in ideas that you have for optimizations for for these different things and then you would work on a seven-day sprint of like how can we increase conversions what tests can we make if running like a discount promotion gonna help is doing a bit email follow-up sequence gonna help its rewriting like the copy on that page it's not going to help is improving website page speeds gonna help customer churn like if we have a better customer like subscription off-boarding flow is that been reducing.

So you should spend like a week working on each of these different metrics so these are like really simple things to to track so you did like I said you don't need to overwhelm yourself of the data that you're that you're collecting because if you just have like a ton you're just sort of like not quite paralysis by analysis but I guess somewhat you're just oh you're just feeding this off too much information and numbers particularly for not getting started and if you're just like a small team instead just feed yourself like data that is really easy to understand there is really significant to your business but these are all really key metrics to any business to any software businesses growth subscription business growth and.

And you should be tracking them and I think only when you get to a point that you feel you're not able to optimize this data anymore which is gonna take you a long time to do that I like in my opinion probably a lot wrong than you think to to like aren't completely optimized like a beat us the hell out of your entire website like landing page and checkout flow signup page and so on and get to a point that you cannot improve it anymore it's gonna it will take a while you're gonna have to run a lot of different tests to get to get to that point and a load of different optimizations and then not to mention that all these different metrics in.

These automations that you could build around these and plug-ins and hacks that you can implement there's at least like a few months worth of work and on a be testing like it should it be something that ever ends you should be like constantly testing but when you feel that you're like collecting this data and you've made all the optimizations that you can based on this then maybe you want to go into like tracking more advanced not advanced but more like intricate details like for example which button is someone clicking to sign up from right.

Now that date is completely useless to it to us really until we've optimized as much as possible on the other metrics that we just spoke about but like then again there's a really good book that I mentioned before called like getting real which talks exactly about this you don't need to track like the overlay and affect the shadow of your like button to increase conversions I guess it's just not worth worrying about to start with get down your core metrics set up good tracking in place and and and and that is what you should focus on.

So now I'm not going to like record the whole process of me doing it but hopefully you understand that the general concept I need to basically sync over from like what I'm just thinking of where I actually want to host this potentially an air table I could actually make analytics eye tracking systems we know the percentage for each of these things how many conversions we yeah now I need to work on making like some connections and syncing over two destinations for example like if a user doesn't doesn't complete conversion we can then go ahead and sync them over to like a custom audience and perform like retargeting we can sync them over to like an email marketing tool I think over on Intercom.

So I now need to like implement all of this stuff and start collecting some data to make some optimizations but I'm really looking forward to it because we have a completely blank slate so we're starting from scratch and I'm in charge of growth on this project so it's my responsibility to set everything here get all of the data that I need and then start implementing some growth hacks to optimizer metrics of course like this will be documented and I don't think anyone's ever done a video like this like they upload videos of sort of growth hacks if before you want to call them I think that's arguable point in a lot of situations but.

Anyway I don't think anyone's ever actually gone through like the entire growth hacking process in terms of tracking metrics understanding what's happening coming up with ideas to optimize these metrics designing them building them implementing them putting in plugins like whatever it may be that you're doing talk to measure that metric then tracking the results and then deploying as a natural fix that is what that growth hacking is actually all about for people with who've read Hacking Growth and things like that they will understand that is what the actual process is.

So yeah I'm really looking forward to doing this and I hope that people will learn something from like the sort of things that we put together and thinking together various different APIs to get our like analytics in place it's getting late now so I'll probably mess around with this for like half an hour or so and then go to bed and like sit down on this properly tomorrow but yeah thank you very much for watching I hope that this has like given some insight and ideas for you on implementing your your tracking and how easy it is to like sync your data between Segment and like various other tools yeah I hope you are.

We can implement as well actually. That's so yeah you can you can actually just get so flexible with this anyway I won't get into another lecture thank you very much for watching I hope everyone is safe and well in this rather crazy situation as always have a great day if you enjoyed the video please yeah please leave a like I would really appreciate it.