At this past SaaSFest, Drip’s Founder and CEO, Rob Walling, went through a deep analysis and case study of how he and the Drip crew used onboarding to effectively increase trial to paid conversions substantially. The best part is all of his advice can be implemented extremely quickly (less than a day to a couple of days of work) and each iteration brought new levels of effectiveness.
What follows is a cleaned up transcript of Rob’s talk, but here are the major points:
Talk to your customers: Your biggest source of what makes your onboarding good or bad are the people who just went through your onboarding. Sure, that seems obvious, but most people either build their onboarding in a vacuum or ask people who went through onboarding long ago.
Find your minimum path to awesome: You need to find out what’s the one (or two) pieces that gets your customer hyped about your product, and then get them there as soon as possible. As Rob notes below, this isn’t always what you think it is either, which makes point one above so important.
Use email and in-app onboarding: Use all the tools in your arsenal to get that trial customer to the minimum path to awesome.
Pre-qualify trial customers: Create a unique experience by asking some pre-qualifying onboarding questions in order to really double down on personalizing the minimum path to awesome.
Keep it simple: Their current onboarding is literally a questionnaire at this point, which dynamically updates with each answer, allowing them to properly keep onboarding as simple as possible, while maintaing the personalized feel.
Starting Drip in a Crowded Marketing Automation Space
A little bit of background on me, and this is kind of the why do you care that I'm saying this? Or why should you listen to me? Or why do I know what I know? I launched Drip, co-founded it about four years ago. We grew it to seven figure ARR, we got acquired by LeadPages about five months ago. And then before that I had a whole suite of apps, a bunch of apps I've started over the years, and probably the most notable from there was Hittail, which was an SEO, keyword tool.
So to set the stage a little bit, we started Drip in 2012, we launched in 2013 in the marketing automation space. So it's a tool that competes with the slide that Brian had earlier. It is similar to, you might think of an Infusionsoft or an ActiveCampaign, more so than a HubSpot. HubSpot is upmarket. We start at about 50 bucks a month and we go up, we have customers paying us a few thousand, at the top end, a month. I used to call it, "Lightweight Marketing Automation that Doesn't Suck," 'cause I really liked that. But once we got acquired, they were like, "Change that," 'cause, I think, "Doesn't Suck," on your homepage isn't professional. But it is what it is.
So we have broadcast emails, we have autoresponder campaigns, as you would expect. We have automation rules, if someone does this then do that, right. You tag them or you move them into a different campaign, or send some email. Then we have a visual builder. We have one of the three or four visual builders available on the market. So that's the gist of the product.
Four versions of onboarding in three years
Now, I'm going to dive in. Again, I have four versions of onboarding that we iterated on over the course of about three, three and a half years, with varying degrees of success. So what I'm going to talk about today is, not only the framework that we used to come up with these versions, 'cause it's the same framework we reapplied as we had new knowledge, but I'm going to show you some tactics as well that we found that worked, and some that we found that really failed kind of miserably.
Onboarding 1.0 – Manual email, manual onboarding
So, this first one is kind of anticlimactic. Onboarding 1.0 is an email I was sending out of my Gmail account when we were manually onboarding people. You're going to get these slides later, so don't feel like you have to read this. I'll read a couple of snippets of it and pull out the salient points. But in essence, if you're early stage, I know most people in here are not, but if you are early stage we onboarded between 30 and 40 of our first customers using my Gmail account and Boomerang. So I basically email out, boomerang it in four days if I hadn't heard from the person. I was, basically, doing a CRM out of Gmail. The interesting part about this email is it actually contains all the steps that someone needed to get started in our app, if they were willing to follow it. And, oh man, this is small. All right.
Onboarding 2.0 – Finding the minimum path to awesome to double trial to paid conversion
But that wasn't obviously going to scale, right? I mean as soon as we emailed that launch list and got 500 people in trying the app, there's no chance I can maintain that. So what we did is we sat down and we said, "We have to find out the MPA," that's minimum path to awesome. It's the minimum path from here to there that gives them a dopamine rush. So, if you have invoicing software, maybe it's when they send their first invoice, maybe it's when they get paid, we don't know yet, we just have to take a guess. If you have proposal software, maybe it's when they send that first proposal. If you have email marketing software it's going to be either when they get their first subscriber, maybe when they send their first email, you don't know yet. You have to take a guess. So you've gotta sit and think, "What is the minimum path to awesome?" Then you have to talk to people about it.
Talk to customers
This is the benefit we had, and everyone in this room has, is v2 of onboarding, we went and asked customers. Because we had them now. And the coolest part is if you ask customers who've onboarded in the past, say, 60 - 90 days, they still remember. If you ask your old customers, "What was it that really gave you that dopamine rush?" No one will remember if they've been there two years. So, you almost have the luxury of if you have people constantly coming in, of being able to figure out what is this minimum path to awesome for a new customer starting today?
So we took that, we sat down very scientifically, we hashed out ... This is the spec for our first thing of onboarding, and it made sense at the time, and this is what we ... yeah, I'll do one more, one more slide. We realized that we, our approach ... This is the framework slide that I'll come back to several times ... The steps are this. One, to determine your users' minimum path to awesome. If you're still in v1 then you need to take a guess, if you have customers then you ask them. Second, you guide new users through it in the app, okay? I'm going to show you how we did that. You can do it however you like. You can borrow from us or you can ... You know, there are some other apps, or you can code it up yourself. But we guide the new users through it in app, then we guide them through it again via email, to pull them back into the app, and then, this is the kicker, we offer to do it for them. This is where we started calling it, "concierge."
Guiding new users through onboarding in-app
Okay. So we'll start, so we've already talked about determining users' MPA, we're going to look at guiding new users through it in app. This is what we did. So this is an early screen of Drip, it doesn't look exactly like this anymore. What you'll notice is along the top there is a black bar with three little, it's like a wizard, right? It's like a progress bar. And a lot of B2C apps get this done really well, as Brian was talking about. The viral loops, they need to get this done with exceptional depth, whereas a lot of B2B SaaS apps don't really focus on this as much.
Once we implemented this, and what I'm going to show you, this is where we ... it was about a two and a half times increase in our trial to paid conversion rate. So you'll notice across the top we just have three sequential things, create your campaign, install our Java script, and then we had an optional define your goal. So these are the same three steps I laid out in that email, they're just done in a way that, when you're in the app, you kinda don't see much else. Now if you clicked on tabs you could do other things. You could go and send a broadcast email, you could do whatever else those tabs allowed you to do, but it was always kinda sitting there bugging you at the top, "Do this," until it got done.
Sometimes you need to do it for them
Interesting kicker is, this is the concierge part, is even when you're guiding people in app, they will still often ignore it if you're adding things to their to do list. So imagine someone clicks, all right, I want to activate a campaign. Well, an email campaign is a sequence of three emails, five emails, right? It's kind of a mini-course as you think about it. So now the customer has to create content. They have to come up with it, they have to pull it from somewhere.
So this is where we realized another sticking point was happening, where we wanted to give them three different options to get through the, "Activating a Campaign," option. So again, if you're invoicing software maybe you give them boiler plate text, if you're proposal software maybe you give them boiler plate text. But we decided we're either going to do it for them, which is the concierge option on the left, we offered it for free. We said if you bring the content, you bring five blog posts, you bring an e-book, we will put together the course for you. Or you can choose a Blueprint, which I'll show you in a second, which is really just like ad libs, you know, it's kind of this whole set out blueprint that you can click, poof, and then it would pre-populate all this stuff and then you'd change some fields. Or, if they wanted to pay 500 bucks, we had a copywriter, essentially, on retainer who would do it for them at cost at 500 bucks. These are the three options and at the very bottom you can start from scratch. If you're an expert, you start from scratch.
This was another game changer for us, where people used to come in and just wander off. This started also forcing them, and they felt guilty if they didn't pick one of these options then at least move forward. So this is where the concierge part comes in. These are our blueprints, really quickly. They're not that sophisticated. You'll see there's just these big asterix and this all caps that says, "Replace this with that." So it wasn't like we built some big, sophisticated system, but if you clicked the button to use this blueprint it didn't, yeah it actually took all the email copy and the delays, right, 'cause these emails are delayed, poof, and it just hydrated right into your account. So it was like this one click thing and you felt like, "I accomplished something today." And then you had to dig in and add the real content. But that was the concierge kind of light bulb moment for us.
Then, with Java Script, 'cause this is always a kicker, getting people to install Java Script's kind of a pain. We took a tact that worked well for us. We had a manual installation, or we added WordPress and Segment. So we gave them as many options as we could to try not to make people have to go write code, basically, or get their developer to go hack into a footer somewhere. So, installing Java Script is always a trick, but if you can integrate it with Segment or you can have a WordPress plug in, which we did, then people can do it without writing any code. 'Cause that's what we're trying to avoid here, right? We're trying to think where's the friction points in getting to the minimum path to awesome, and how do we think about every way possible to eliminate those as creatively as possible.
With a previous app, Hittail, I talked about, we actually had a button on this page which said, "Click here. Email us. We will install us for you if you give us your ..." whatever, login, FTP access, or something. We would go in and install it in the footers of their site. We decided not to do it with Drip, but that worked well for us as well.
Tell them via email, too.
And then, this is the second part. So remember, I said figure out the minimum path to awesome, do it in app, then tell them again via email. This is the tell them again via email part. We have just a five ... I think we still use this course almost identical today, with this sequence. It's five emails that went out to, the go out to our trial users. You'll see that open rates on these things are shocking, 68%, even the bottom one's 59%. We really get high open rates on it. The text of these emails ... They're sent out over a 21 day trial, and the text is something similar to this, "Drip's goal is to help you make more money," blah, blah, blah. We want to make it ludicrously simple to collect subscribers. "Your free trial is x days long, and these are the three steps to get set up." So it's the same thing. I'm telling you the same three steps. It's activate a campaign, install Java Script, and create a conversion goal.
Then you'll notice we have liquid templating in here that is nice. If the subscriber tags contains, "Completed onboarding," then we do blah de blah. So it's a little bit of dynamic stuff in there. Actually you don't need, need, need that. It'll still work without that. But this is a nice perk of using an automation platform to do it. So you could code this, send this straight out of your app using SendGrid Transactional Email, or you could use a platform like an ActiveCampaign, Drip or InfusionSoft. So again, this, to be honest, this single approach when I acquired Hittail in 2011 and they weren't sending any onboarding emails. And this is really, this is five years ago, this is before a lot of people were talking about doing this kind of thing, and I implemented a sequence like this. This was the one that tripled, almost overnight, tripled that trial to paid conversion rate on Hittail.
So again, our approach was determine a users minimum path to awesome, you guide them through it in app, you do it again via email, which is simple. And the via email part has as much, or more of an impact as the in app part, and it's a hell of a lot easier to implement, because writing the emails and having them just sent out through an auto responder is a lot less work than reorganizing your app and adding the top toolbar and stuff. That's the ticker. That's the one thing I would do next week, if I were you, if I were thinking, "What do I want to implement first?" I'd do the emails. Okay, and then offer to do it for them, concierge, try to remove all the friction.
So this, by the time we launched and had all this in place, had our trial to paid conversion rate, with credit card upfront, at between 50 and 60 percent, which is where it has stayed since then, and we were happy with that.
Onboarding 3.0 – Custom onboarding flows for each type of buyer
Okay. So that was onboarding 2.0, and that was essentially the day we launched in 2013. Then, what we found, so this was the thing. You know how Brian was talking about your marketing approaches they're just going to get their ass kicked, and then you have to figure out a new one? Your onboarding is going to do the same thing. About every year we've found we've had to redo our onboarding, which is a bummer, because you kinda want to set it and forget it, and you can, for about a year, at least based on our experience. What we found is we built a lot of integrations with Drip. We use this approach called, I call it integration marketing, which is where you're trying to get as many people tied into you as you can, trying to get as many customers to use it as you can, and, like most of you in the room who've looked at the data, if you get your customers tying your app into someone else's, you're going to have higher retention. We found that across the board.
So we were trying to figure out, you know, nowhere in that minimum path to awesome, and in any of the onboarding that I just showed you, did we talk about integrations at all. And so, after about a year we had 35, 40 integrations with things like, I would say Stripe, and Paypal, Shopify, they could easily pipe right into your Drip account, or landing page providers OnBalance, LeadPages, Calendly, et cetera.
And so what we realized is, we were going to try to integrate it into the minimum path to awesome, but we realized that if we just asked a few questions up front, we added just two or three screens before the onboarding, that we could actually get a lot of information out, and that almost everyone, it was like a 90% completion rate because all it was were questions. No one had to do any work. It was like a survey, and they felt like they were accomplishing something just by answering questions like this, "What do you want to do first?" This is the first thing. You enter your credit card, you click, "Start a trial," and this is the first thing you saw in Drip. We'd say, "Hey, what do you want to do first? You can always change this later. Are you sending marketing emails? Are you sending emails to your existing customers? Are you doing digital products? Or do you have a subscription service?" Then we started narrowing down. So if you said subscription service, then we know Stripe, Recurly and Chargify at the time were our three subscription providers. I think we have a couple other since then. Or if you were selling digital products, all we had at the time was Stripe and Gumroad, we have a bunch of others now.
But the idea was to figure out what are they using, and then to add that into the onboarding itself, and then make that a step. Like, if you're using Stripe, then let's auth you in there, and get that tied in. 'Cause as soon as someone saw their new Strip subscribers, or their new Stripe customers being piped into Drip with a customer tag, then you have their information. It's like, "Oh, that's pretty cool," and so it's a nice dopamine rush. And so we did. You'll see at the top that it says, "Install your Java Script," and the second step now is, "Connect to your Stripe account." And this is where we got fancy, and started saying, "We're going to have the top nav shift based on what they said on that earlier survey." So this is a v3 or v2, right? This isn't something you do the day you start. But this is working pretty well too. This started ... This not only got, it got about the same amount of people onboarded, but we saw more people stuck around. It helped our retention long term throughout the cohorts, okay, so this is all good stuff.
Onboarding 4.0 – Breaking things down to be as simple and personal as possible
Then we launched Workflows. In January of this year we launched this visual work flow builder, which was a huge, it was a complete game changer for our company. It was one of the reasons we got acquired, and it's one of the reasons that ... it accelerated our growth dramatically, almost doubled our growth month over month once we launched this, 'cause it just made us a player, a major player in this space. So this is January 2016, and I've made it a habit since we are boot strapped and we're in control of our destiny, we ... I never committed publicly to deadlines, 'cause you never wanna say, "Hey, I'm going to launch on 27th," and then you find you're working 90 hour weeks to hit it. So we've never done this except this one time, literally this one time. On 13th January we sent an email, and it said, "Hey. We're going to be launching something really cool ...," and we had this kind of blurry vision of the work flow. We were hinting that we were going to launch a visual builder, and our customers were all in a froth, and there were all these comments and stuff. It was really cool.
And then the next day, on 14th, my co-founder Derek said, "What are we going to do with our onboarding? WorkFlows is the app now." The whole app became this visual builder. We still had all the other stuff, but it was so much less important. But if you look back at our onboarding, all it does is lead someone to install Java Script, activate a campaign, and set up a goal. It had nothing to do with what the app was going to be in two weeks, in essence. So this was onboarding 4.0. We were kind of backed into a corner, and this is the one that we built in essence in two days, because we had no other choice.
What we did is, we went into a conference room and we took about 90 minutes, and we said, "What if we rip out everything we've built to date? We literally rip out all of that top nav, we rip out all of the pre questions that we've built over the last couple of years, we keep emails, 'cause the emails work so good and there's no reason to change them. What if we simplify it, and we basically make it one page? Maybe two?" And so that's what we did. First we realized that we were missing out not asking questions like this, where if you can ask a first question where you can figure out where someone's going to be in your pricing tiers, which is what subscribers are to us, that dictates how much money they're going to pay us, very, very, helpful. So this was just a no brainer to do. We get 85 or 90 percent fill on that phone number field which is shocking to me. We say we will only use it for billing purposes, which we do. We wouldn't call them for other reasons. But this was a big revelation for us.
Then this is the one page, this is really the one page. We took everything I've said so far, and we put it into this little kind of wizard, and people love it. We literally get people tweeting at us. When we were being acquired we had two people in LeadPages just tell us this is the best onboarding experience in a SaaS app they've ever seen, and all it is is a nice, well built wizard, where if you click, "Subscription Services," then it puts the right thing below it, and it just kind of ... Again, it's a questionnaire, and it makes you feel like you're getting something done while you're doing it, as a customer.
So we ask, "Do you use a landing page provider? Would you like to use Drip to market products or services? What type of product do you sell?" And then, "Payment provider," and, "Do you offer a free trial?" And this helps us, basically, build a custom onboarding sequence. So instead of having the three buttons across the top, we actually throw them into a work flow now, and on the upper left you'll see that's the to do, that's the onboarding, that's the new nag. Aside from the email, is that thing in the upper left saying, "Before starting your WorkFlow ...," blah de blah de blah. So this has worked really well for us.
What I want you to take away from this, obviously you don't have work flows in your app, but the concepts to take away are number one, if you're going to do one thing, do the emails. That's going to be the most important. That's the one that I've seen game change the most apps. Number two, if you have a bunch of integrations, and you're not surfacing them somewhere early in the process to get people to wire up, especially if they're nice integrations, people can just click a button and then stuff magically happens. Try to think of a way to do that, whether you do that first page approach that we took, or whether you do something else, it's a big deal.
What was consistent for all versions of our onboarding? We had four versions that we went through, all the way from an email to what you saw on the last screen, which is using in app, a way to get people onboarded. Determine MPA, you guide them through it in app, you go it again via email ... oops, skipped to many. I'm going to leave this slide up, I think. You do it again via email, then you offer to do it. A lot of people poo-poo the concierge stuff, because I've heard, "Well you can't do that at scale." It doesn't matter if you can do it at scale, you know. We did it up until we were a couple of thousand customers, and we were getting several hundred trials a month, say 500, 600 trials a month. We were still able to keep up with the concierge load. It was one support guy, who I trained to take in content and spit out the course. So if he spent an hour on it, and it was 15, 20, 30 bucks and hour, whatever it is with burden, it was worth keeping those customers.