Nivi · February 17th, 2010
Thanks to KISSmetrics for supporting our interview with Sean Ellis. If you want an intro to KISSmetrics, send me an email. I’ll put you in touch if there’s a fit. Thanks. – Nivi
Hiten Shah from KISSmetrics recently sat down with me to discuss how to optimize funnels with their upcoming analytics tool, KISSmetrics.
You may know Hiten from his Crazy Egg days and survey.io, which Hiten and I discussed in How to measure product/market fit with survey.io.
SlideShare: How to optimize your web apps with KISSmetrics
Audio: Interview with chapters (for iPod, iPhone, iTunes)
Audio: Interview without chapters (MP3, play anywhere)
Transcript: See below
Prerequisites
You’ll get more out of this interview if you also read:
- Our interview with Sean Ellis parts 1 and 2.
- Our previous interview with Hiten: How to measure product/market fit with survey.io.
Outline
Here’s an outline and transcript of the interview. The interview and transcript are about 23 minutes long and we’ve highlighted some of the juicier bits for you.
- After fit, prepare for growth
- KISSmetrics helps you optimize your funnel
- KISSmetrics helps with all optimization steps
- First user experience isn’t necessarily the paid user experience
- KISSmetrics is most valuable after fit
- How Cloudfire uses KISSmetrics during fit
- Startups aren’t a science — but we’re getting closer
- How other people use KISSmetrics
- KISSmetrics is on it’s third iteration
- Why KISSmetrics’ hasn’t launched
- KISSmetrics tracks actions on a per-user basis
- KISSmetrics lets you calculate customer LTV
- Get more startup advice
Transcript
Music: Squarepusher
Trent Wolodko: Hey there everybody, my name’s Trent and I’m new to Venture Hacks. I’m going to be helping with producing content for the site so that the guys can concentrate on bringing you good advice faster and more furiously than ever. Today Nivi speaks once again with Hiten Shah from KISSmetrics.com. He spoke with Hiten recently about survey.io and today he’ll be talking to him about optimizing your web apps using the tools over at KISSmetrics. Right off the bat, these guys make a lot of references to Nivi’s interview recently with Sean Ellis about bringing your product to market as well as his former interview with Hiten regarding survey.io. So, you may want to check out those interviews as well on the site. The guys have a lot of ground to cover as you can see, so without further adieu, I’ll throw it over to them…
Babak Nivi: So the second part of the interview with Sean is what activities you do to prepare for growth. Right?
Hiten Shah: Yeah.
After fit, prepare for growth
Nivi: One of the insights that I got from the interview was that you do a whole bunch of stuff to get to fit. That’s step one. Then step two is that you do a whole bunch of stuff to get ready to prepare for growth.
Hiten: Yeah.
Nivi: And Survey.io is great for the product/market fit stage, where you’re trying to get the fit. And you guys also have a product called KISSmetrics, which helps you prepare for growth, or to put it another way, to optimize.
Hiten: Yup, and that’s our main product.
Nivi: So, what does it do?
KISSmetrics helps you optimize your funnel
Hiten: First, it’s an analytics product, pure and simple, but the difference is that it’s meant to do one thing really well, and the thing it does really well is help you optimize a funnel – help you understand what your funnel looks like for whatever type of product you have.
Nivi: So, Sean has a few steps in the optimization stage, right?
Hiten: Absolutely.
KISSmetrics helps with all optimization steps
Nivi: Economics: put into place a business model and optimize it. For example, change your price or change the way that you charge, completely, whether it’s by the year or by the pound or however. So that’s the economic step. That requires optimization.
There is the positioning step – how you describe the product – which also requires optimization.
And then there is the step of essentially moving people through the funnel and getting them to the finish line. That requires optimization.
Hiten: Yup!
Nivi: How do I use KISSmetrics in any of those?
Hiten: You can use them in all of them, because when you have economics in place it means that you’re charging customers and that you need to understand things like how different price points affect your revenue. In essence, it’s a funnel. You’re still funneling people through a sign-up process, and you might be A/B testing two different things.
So we facilitate some level of A/B testing today, because you can pass in the data and visualize it, but in the future we’re going to start adding more components around A/B testing and optimization. Right now it’s a hard focus on helping people understand their funnel, but then that goes into both of those areas we were talking about.
So, in the economic stage it’s just about when you’re testing different pricing plans and things like that, and understanding if there are any differences, also, in different channels of customers depending on where they came from, and if your revenue is different for those based on the price points.
Then on the promise side you’re probably doing a lot of things around testing landing pages and things like that, so you’d build different funnels for different landing pages and different cohorts of users, and we completely facilitate that, as well.
Nivi: So you let me – and maybe you can just send me a couple screen shots that I can put in – compare two cohorts, how many people get to the finish line based on their channel, based on the positioning that they saw on the landing page, based on the final price or business model that they saw that we offered them, and also based on what Sean calls that “first experience.”
Hiten: Yeah, first user experience.
First user experience isn’t necessarily the paid user experience
Nivi: The first user experience, which is just kind of the funneling. I mean, it’s everything.
Hiten: Yeah, but it’s primarily how to optimize your on-boarding process of first-time customers so you can make that more efficient – efficient meaning highest conversion rates – and so that you can spend on marketing and you understand what the value is of each user.
Nivi: What does that mean to you, other than a conglomeration of the landing page and all the steps between that and putting the credit card in or signing up?
Hiten: In every product there are a number of steps people take to get gratified with the product. Gratification, usually in freemium products, happens during the free period, most likely, that instant gratification, and then they have another reason to go pay. So it’s all about understanding how people are moving through those stages of their usage of your product. So the idea is to track all the different events of gratification and optimize against that. And there are always steps in between before they’re gratified.
For the first gratification, they have to sign up. They probably have to fill out a form. They probably have to install something, or depending on what type of product it is there are a number of things they need to do before they feel satisfied that this is a must-have product. And once they hit that point, then you’re going to have an ask of asking them to upgrade, and making sure they hit a point where…. It’s really about you trying to understand what point they hit before they’re willing to give you money, and understanding the different things you can do around that to encourage them to give you money.
This all has to do with, if you have free plans, you’ve optimized your first user experience. Now you really need to understand what the prompts are and where you put them to get them to pay you money. So it’s all about prompt optimization.
An example is if you have an analytics product and you have a free plan, but you want to charge users, you probably need to put prompts in different places, as customers do different things with the product, to get them to pay, basically.
Nivi: In order for KISSmetrics to be useful to me, should I be done with the product/market fit stage. In other words, should I have 40% must-have users, or can I use KISSmetrics earlier, effectively?
KISSmetrics is most valuable after fit
Hiten: The short answer is that I think KISSmetrics is most valuable after product/market fit. But, this is one of those things where creating a startup and the steps to creating a startup aren’t defined yet. People look at them in all different ways.
I think both Sean Ellis and Eric Ries are making a lot of noise and movement, and there are a few other people like Dave McClure that are doing all of this, to get people to think of it more as a methodology, and there are different stages to a company and different things you should do at each stage. So if you follow one of those methodologies or one of those mindsets or theories, what you really want to do is wait until you have product/market fit before you optimize first user experience.
My own example is that we’re not optimizing our first user experience for KISSmetrics right now. We’re getting users in, but we’re not optimizing it because, really, that’s not the important part right now. The important part for us is getting enough users and building and iterating on the product to make it so that it’s something they can’t live without. And that’s not a quantitative thing; that’s not something where if I can measure how they go through a funnel, it’s going to necessarily help them. It’s more about things I need to add in the product, and it’s a little more of a fuzzy goal. Honestly, hitting product/market fit, in my opinion, is a pretty fuzzy goal.
Nivi: Tell me how Ash, who has written about this…What’s the name of his company? CloudFire? Is that right?
Hiten: Yeah.
Nivi: How is he using KISSmetrics actually in the product/market fit stage? Is that right?
How Cloudfire uses KISSmetrics during fit
Hiten: Yeah. What he’s actually doing is combining a few ideas. Dave McClure has something called AARRR!! that’s a framework that he came up with years ago. He’s also one of our advisors, for a good reason. And he spread a lot of the religion of metrics, so a lot of people say he’s the church of metrics.
But basically, what it does is break down your business metrics for a startup into 5 different things.
Acquisition, which is metrics around how you acquire customers, what channels are profitable or what channels are bringing in customers and how you’re bringing them in. So it’s more like market.
Nivi: So you define acquisition.
Hiten: Yeah.
Nivi: One definition might be: they signed up.
Hiten: Yes, exactly. They signed up, and where they signed up from.
What Ash has done is taken acquisition and made a funnel for it, and he calls it the acquisition funnel in KISSmetrics.
Then the next one is Activation, which is how to get them to have their first gratifying experience. They’ve activated with the product. They’ve used it. They’ve experienced it.
Nivi: And in his case it’s something like, they downloaded my software and uploaded some photos with it. Something like that.
Hiten: Exactly! Exactly. It’s like they activate it, which means they downloaded it and uploaded a photo. They had the first experience of using it.
And then the next one is Retention, which is that they came back or they used it multiple times or some factor of that. It depends on what your startup is. So for him it may be that they used it once, but then they used it a second time and that would be retention.
Nivi: In thirty days, say, they use it again.
Hiten: Yeah. So he’s got another funnel around that.
And then the next one is Referral. That one is more like, are they referring other users? And that’s a whole thing in itself; it’s ‘forward to a friend’ and ‘contact importer’ kind of stuff. There are all kinds of things around referral.
Nivi: Tweet about it.
Hiten: Yeah. That also gets into virality and all of that kind of stuff.
And then the last one is Revenue. So he’s got some revenue funnels around all of the prompts. I don’t know if this is exactly what he’s doing, but this is what I would imagine he would be doing. I don’t look into his account; I just talk to him and read his blog posts.
Nivi: That’s good.
Hiten: He’s probably got a funnel that’s focused around the prompts that he’s using to get people to pay. So he’s actually using it [as an] analytics tool, in general. People end up using it whatever way they want, and that’s one of our challenges, which is that we really need to figure out what type of use-cases we want to support.
Nivi: Question number 6 or 7.
[laughter]
Hiten: Yeah. You got it. So we’re not even there yet of giving out that survey, but we will be soon enough, hopefully.
Nivi: I think it will be interesting to see how his activation and retention numbers tie into the must-have segment.
Hiten: Agreed. Yeah.
Nivi: Probably mostly on the retention, if I’m thinking about it right now correctly, because for the person to even be a user that you would survey with Survey.io, you’d probably want them to already be activated.
Hiten: You might even want them to be a retained user, and hit the retention step.
Nivi: Yeah, you’d want them to be retained. Right.
Startups aren’t a science — but we’re getting closer
Hiten: So I think he’s actually using it in a very interesting way that we’re probably going to study a little bit deeper, because Dave’s an advisor and we buy into his theories. Actually, our first iterations of this product, which not many people saw, each of those was a view in our system, so we’re actually very intrigued by what he’s doing with our product. But again, I don’t think startups are a science yet, and I think we’re getting closer and closer to it, so it just really depends on what people subscribe to. All we can provide is a tool at the moment, and then hone in on these different use cases of it.
Nivi: Right. One thing that Sean actually mentioned in the interview is that he doesn’t see a lot of correlation between the recommendation figure and “success of the startup.”
Hiten: Right.
Nivi: Whereas he sees a fair amount of correlation between the success and the must-have figure. So that’s one thing to keep in mind for Ash, perhaps.
Hiten: Sure.
Nivi: And then, too, just in your case, like you were saying, with KISSmetrics, the way you’re bringing it to market is very feature oriented right now, and then you’ll figure out the positioning based on qualitative conversations with customers.
Hiten: You’ve got it. Right now it’s more about product and how to tweak the product. I hate to call it features because for us, actually in the market we’re in, every analytics product has hundreds of features. We actually want to provide the features that matter, so that’s kind of the difference.
It’s basically iterating on the product and making it so people like Ash love it. There are probably a number of things… He might already love it. He might be a very unique case where he’s been looking for something like this. But a lot of other people, I get on calls with these people or talk to them, and….
Nivi: So how do other people use it?
How other people use KISSmetrics
Hiten: How they’re using it is that there are several Facebook applications that are using it. They’re using it to optimize their own funnels and just optimize the conversion rates between each step, and things like that – normally what you would do with funnel optimization.
And then other folks, like SaaS businesses that are using it, are buying traffic from AdWords and they’re passing in the dollar value for their customers at one of the steps, and they’re just basically optimizing the funnel, but based on revenue because that’s what their main goal is and they’ve already got their revenue set up.
Nivi: So, visualize the funnel and visualize cohorts, and compare them.
Hiten: Yeah. And some of this stuff is still things we need to do better in our own product. We’re very early. We released this version on November 1st, so we’re not even a month-and-a-half into it.
Nivi: You guys have basically realigned the company around this product, is that right?
Hiten: We’ve realigned the company around this product through doing our own customer development this whole time – “this whole time” meaning this is our third iteration of the product and this is the one that we’re being a lot more public about.
Nivi: What were the first two?
Hiten: With our first one we had assumptions around doing metrics for Facebook applications and social network applications. And what we realized there was that we had a very constrained view, and it was based on Dave McClure’s AARRR!! stuff that we just talked about. What we realized about that was that people started asking all kinds of questions that our interface could not answer for them.
So we took what we learned there and we built a second product, which was very super-flexible event tracking, and we had no funnel analytics or anything like that, so this kind of stuff would help you with tracking engagement of users. If you were a game it would help you track game plays by gender and all of those kinds of things. And if you were a SaaS app, it was more about if Ash needed to measure what types of photos people were uploading or what kinds of things they were storing on his system, we would provide metrics around that kind of stuff.
Nivi: And what was that called?
Hiten: It was KISSmetrics. All of these are KISSmetrics, we just didn’t publicly release them.
Nivi: OK, got it!
Hiten: And so we used these other two products to kind of learn about the market and figure out what our target should be. And it was super-generic products, especially the second one. The first one was very targeted towards a market, then we kind of ran away from that and said, let’s get Facebook apps, but let’s also get SaaS applications; let’s get blogs using our product. And that was a very highly flexible product.
In some ways it would have been really fun to market that product and deal with it, but we were just heading down a path, in terms of sales and marketing and things like that, that we didn’t want to. As a business our DNA was more around building a really simple solution.
We have a previous analytics tool that the people in our company built called Crazy Egg, and it’s a simple tool, and we wanted to go back to that thinking and say, analytics is something where people log into analytics systems, and a lot of the time if they’re not expert users they’re very confused about what they’re looking at, and then they end up just using it as a trending tool just to see general trends.
And the kinds of features that we felt were most important were conversion goals, conversion rates and funnels. So with all this learning from our previous products we basically decided that we should hone in on that proposition, and instead of tacking funnels onto an analytics product, like the other one, we would just make that the product. And it was a big bet. I’m actually super excited about it and very happy that we went in this direction because I think we can really make an impact on the market and produce something really useful for customers.
Nivi: The feature is the product, here.
Hiten: The feature is the product, and that’s a trend. I mean, Twitter is just a feature, right?
Nivi: Right.
Why KISSmetrics’ hasn’t launched
Hiten: A lot of people have said, “Oh, you guys are creating a lot of buzz. We know your brand, but what the hell is your product, and when are you going to launch?” And for us it’s not about that. It’s not about launch; it’s not about things like that. It’s about building a product that people love and iterating until you get that.
Nivi: Yeah, and this is a good example of why you probably don’t want to launch until the business is working.
Hiten: Yeah, exactly. We don’t even have our economics down, so we’re not charging anyone right now, and everyone can come in and use the product. At some point we want them to help us figure out what the value is and what we can charge for it.
The way most analytics packages charge, people are penalized for being successful, and we want to see if we can change that a little bit. I don’t know what it means yet. I don’t know what it means for our pricing model, but I’d love to make it so that if you get more traffic we don’t necessarily charge you more. It sounds really fucked up and weird, but yeah. Excuse my language, but we want a better pricing model if we can find it, that’s more aligned with our customers. So there might be more features that they pay for, and it might be access to data or something that’s not exactly like: you get more paid views or you get more growth and we charge you more money, and it scales up like that. Even for us, that’s not necessarily the most scalable business model.
Nivi: Sean talked about using action tags with AdWords, in the interview, to track users on a per-user basis. We talked about Google Analytics, which doesn’t really track users on a per-user basis. What does KISSmetrics do differently here?
KISSmetrics tracks actions on a per-user basis
Hiten: The things we’re trying to solve are if someone comes to the site multiple times and then they sign up, let’s say on the third or fourth time, we’re allowing you to identify that user and we’re trying to attach all of those times that they came, previously, to that customer. What that means is that we’ll understand the first touch point they had with you and where it came from, and we’ll be able to attach that for the history of that user using your product. And that’s a much different view than most analytics have, which is either page-view tracking or event tracking. This is actually, honestly, people tracking, which means that we’re trying to actually track people.
Eric Ries has a great quote on this which is, “Metrics are people, too,” and we firmly believe that. If we can track things on an individual-user basis without getting into a lot of privacy concerns, because you’re already doing this in your data base, so you just have to have the proper terms-of-use for it, we can provide a much more valuable analysis on your user base. And as long as we’re not doing anything in aggregate there’s not any real privacy implication, while with Google Analytics I feel like it’s very hard for them to do that because they have a huge business that has nothing to do with analytics, in a lot of ways.
Nivi: Right. Does tracking people on a per-user basis let you calculate LTV, and to put it another way, cost of customer acquisition, and does Google Analytics not let you do that?
KISSmetrics lets you calculate customer LTV
Hiten: I’m not going to say that Google Analytics does not let you do that. We definitely think there’s a lot of value in the product, and we actually use Google Analytics in a lot of our products, and they do different things than we do. But legally, you cannot pass personally identifiable information into Google Analytics. So, if you’ve got 100 conversions in Google Analytics, you might have 110 in your data base simply because you’re not able to identify each individual user and they might lose some of the data.
And they’ve got, my guess is, 100 million sites using the product. That’s a high load. We’re probably never going to have 100 million sites using our product. There are a lot of issues that come around dealing with that many customers with a free product.
Nivi: Yeah. What I’m just trying to understand is what the requirements are to calculate cost of customer acquisition. Have you guys thought about that at all?
Hiten: Yeah, absolutely. We want to make it as simple as you passing a dollar value every time an event occurs where someone’s paying. And then in our interface, it’s not like magic, but it’s magic. We can show you revenue by referrer, revenue by channel, revenue by plan type, and all of those kinds of things that typically you would have to do a lot of data base queries in your own data base, your user data base, to get all this data out.
And honestly, this is some of the pains we had with our previous products. So, for the first 6 months, we never tagged referrer for each customer that came in, so we couldn’t boil it down to what referrers were creating the most revenue for us. Once we started doing that it was enlightening. So KISSmetrics is actually based on a lot of the pains that we had with our previous SaaS product.
Nivi: So the dollar value that you pass to KISSmetrics will let you calculate the LTV, to some degree, of the customer, right?
Hiten: Yeah, that’s the goal.
Nivi: But not on the cost side. It doesn’t tell you how much it cost you to get the guy.
Hiten: Not today.
Nivi: Right.
Hiten: That’s a really good question, and that’s something that we do want to answer in the future. Obviously it depends on what type of customers we get and how they want to pass in the data. Is it automatic? Is it manual? Things like that. But it’s a problem that we’re thinking about a lot because then you can have a full picture in one dashboard. It’s all about that one report to rule them all, which doesn’t exist in analytics today. And if it does exist, it’s an internal report that’s built by your development team, and that’s kind of the problem we’re trying to solve.
Nivi: Cool! Thanks again for spending the time talking and also sponsoring the interview with Sean and making it free for everyone.
Hiten: Absolutely! Our pleasure.
Get more startup advice
Trent: You’ve been listening to Hiten Shah, from KISSmetrics.com on how to optimize your web apps. And if you want to find out more, you can check out the KISSmetrics.com website. You can also follow Hiten on Twitter @hnshah, or check out his personal site at hitenshah.name.
Be sure to also have a listen to Nivi’s interview with Hiten regarding Survey.io, and the two-part interview with Sean Ellis on bringing your product to market, on the VentureHacks.com website.
Of course, if you’re into all other sorts of good advice for startups, head on over to Venture Hacks and just soak it all in.
Thanks again for stopping by. See you next time!
Topics Customer Development · Interview · Podcast · Sponsor