Pear Analytics

The following post outlines a very useful hack for KISS Insights that allows you to run a “multi-variate survey mode” where you can evenly rotate surveys, which is not currently available in their software.  With Pear being a “power user” of KISS Insights, they graciously helped us build this, and so we figured we would share it with the community.  Enjoy!

KISS Insights is a powerful page level survey tool that we started to use heavily to find out what people really wanted out of our software, as well as real-time direction from users as to what features we should build next.

But we needed more. We actually wanted a way to test pricing where we could evenly rotate a pricing question on the same page within the software – kind of like multi-variate testing. We wanted to know if users would pay $10, $20, $30, $49 or $99 per month for the service. By getting an equal amount of responses for each, we can use that data to come out of the gate with the most widely accepted price plan (if you’re wondering, as it turns out, it was NOT the $10 price point).

The limitation in KISS Insights was that we could only test one question at a time per page, which meant we would have to set one survey at one price, let it run until we had 30-40 results, and then go in and deactivate that survey, and activate a new one with a different price point. Not only was this kind of cumbersome in the UI, but it was going to take weeks, or even months to get an answer.

So internally, we figured that what we would do is hack KISS by creating 5 or 6 Insights accounts, and then using tracking cookies, assign each unique visitor a different code snippet from one of the accounts, and each KISS account would have a different pricing question.  Didn’t seem that hard to do, and logically, seemed like it would work.

So I contacted Hiten and asked him if he thought this was a good idea, and he actually responded with something much better.

In a few days, KISS actually developed some custom code for us that we could attach to our existing script to enable the multi-variate survey mode we were trying to create ourselves. All we had to do was create each survey, then get the survey ID and replace it in the additional code snippet:

_kiq.push(function(){
if(KI.location.matches(‘/free-seo-analysis(/.*)?’)) {
var surveys = [XXXX, XXXX]; // Replace with ID’s
var ab = new $KI.Cookie(‘ab_test_1′, surveys[Math.floor(Math.random() * surveys.length)]);
if(ab) {
KI.show_by_id(‘survey’, ab.get());
}
}
});

So our entire code snippet looks like this:

We found that the “if(KI.location.matches(‘/free-seo-analysis(/.*)?’)) {” line to be optional, and the snippet will work without it, although it does force a URL match, allowing you to use whatever URL’s when setting up the survey.  Now, the 5 surveys are rotating evenly each time the page is viewed, and I’m testing 5 different prices at the same time, allowing us to get an answer much faster.  And as I mentioned in the beginning, the answer was not the lowest price, but rather the mid-level price.

Thanks to Hiten and the KISS team for making this so easy!

5 Comments »Tags: kiss insights

I’m going to show you how I concepted and built not one, but two recurring revenue products in one evening using Unbounce, Wufoo and Chargify – and I never wrote a single line of code.  And the results were unreal.

I am not a programmer.  I wish I was, but I took a different path after engineering school and went more into the marketing side.  My programmers are busy.  I can’t bother them, but I need revenue.  Do I get them off what they’re doing, or can I build something myself to test with some users?  There has to be a way….we get plenty of traffic, and somehow I’ve got to get the 1,500 people a week using our software, and the 10,000+ users in our mailing list to check this out.  I’ve got to do something without bothering engineering just yet….just till I can prove this works.

OK, here we go.  My first product is simply an SEO copywriting service that I just want to set up as a one-time fee for now.  $35 per article sounds good, and we’ll even write the tags you need (title, meta description, H1, etc.).  My second product is a link building service, and I want to get folks on a small, manageable plan, but one where we can keep working on month-to-month, chipping away and building good quality links the old-fashioned way – manually.  Both of these sound totally un-scalable and an absolute nightmare to execute.  But that’s OK because we’re engineers and we can figure out the process flow because that’s what we do…make things efficient.  (And I want to do a blog post later about Software AND a Service, versus Software AS a Service).  Why do I want to do this? Well, we learned from our software, and the 20,000+ websites we’ve run and the tons of customer feedback, that they need MORE than someone telling them their site sucks.  They want you to fix it.  I mean, you don’t go to a doctor and he tells you everything that’s wrong with you, and then send you on your way with a “good luck”, right?  There HAS to be something in adding a reasonable and reputable service to the software we have.  Yeah, it’s going to take some work internally, I mean this isn’t all “hands off” – but hey, it’s really no different than staffing up a bunch of support people anyway that most SaaS products are destined to have.

The first thing I did was log into my Unbounce account and set up my landing pages.  This beautiful interface lets me build landing pages in a snap without needing a single programmer.  The link building page took about 30 minutes longer because I had to spend a while playing with buttons, colors of the boxes, etc.  In the end, I thought it turned out OK.  You can see the SEO copywriting page here, and the link building page here.  Now you can see the modified page after tons of iterations, and lots of help from the guys at Unbounce

The second thing I went into my Wufoo online form builder account and created some info gathering forms in a jiffy.  Once a user clicks on the call-to-action button from the Unbounce landing page, they get dumped into the form and I ask them all of the info I need to know in order to start the job, right, so it’s like a “job request form”.  Easy.  Then I went back to the Unbounce landing pages and linked them all up to the various buttons, and I did send each unique landing page to a unique form, because depending on what they clicked on originally, I needed to send them to a specific Chargify page.

Now, enter Chargify, the leader in recurring billing for web 2.0 services.  These guys are really awesome, because they let me create all the products I want, and they only start charging me after I’ve hit 50 customers (regular charges from your merchant, like Authorize.net still apply).  So I created a product family for “SEO Copywriting” and one for “Link Building Services”, and under the link building services, I created three different products based on the three monthly plans we’re offering.  They let me easily create a Payment Page that all I have to do is link to from the Wufoo form after they hit “submit” on the form, and they’ll send them on over for payment.  Oh yeah, I also used a KISS Insight page level survey on each Unbounce landing page, so I could gauge weather I was full of shit or not.

The result? Well, after an email blast this morning and linking to these pages in our existing software (which I can also do myself), we got 600 leads in the first day, and $1,450 of recurring monthly revenue and $350 of one-time revenue since 11 am this morning.

I get a confirmation from Wufoo and Chargify when someone completes the forms, so then all we have to do is match them up afterward, and start working on the project.  Chargify takes care of the auto-billing for me, and now all I need to do is let people know we have the service now (a marketing function, not an engineering function).

As for rapid iteration, I’ve already created a new product in Chargify and a different Wufoo form for 4 SEO articles per month for $100, and for $150 we will install it on your WordPress blog for you, and already we’ve gotten some folks to sign up.  Took me 8 minutes to create the new product.

At the end of the day, I completely hacked all of this together, but I didn’t need to write code, and I didn’t need to bother a code writer.  I can test various products and pricing, and let the engineers focus on something else.  My next task is to make sure we can keep a high quality product, fast turnaround (some of the articles are already done), and scalability.  Who knows, next week we might try a few more products

Update 9/22/10 2:31p – I just set up Chargify to send a successful transaction to a “thank you” page I set up.  Turns out I got 3 chats today asking me “what should I expect”, so now they know   I also set up a B page in Unbounce for the copywriting service to sell a 4 articles for $100/mo plan.  30% better conversion rate on that one!

Update 9/23/10 4:58p – I just hacked together a very simple rating system, again, no code required.  Since we started sending articles back to folks, I wanted to see what they thought of it, so in the bottom of the email, we put “Please Rate this Article: Awesome! |  Good :/ |  Sucked ” which then takes them to a page I created on our WordPress blog which has more information about how to fix an article, have us re-write it, or even order more.  My thought was I can track unique pageviews to identify a trend or problem.  It’s probably not the most effective thing in the world, but it works for now.

Update 9/24/9:58a – Yesterday I used Unbounce to create a variant of my Link Building landing page, which was sucking wind.  For whatever reasons, people think link building is a “spammy” SEO tactic.  So I went in and created a new variant to try and address that, and BAM!  1,271% better conversion rate!!

47 Comments »Tags: chargify, kiss insights, link building, seo copywriting, unbounce, wufoo

There are two main problems being a startup: 1) we have limited capital which means we can’t waste money on poorly performing campaigns (which is a good thing, actually); and 2) as founders we have to do everything ourselves and learn all of this marketing stuff, and its usually the hard way.

So, here are 4 easy ways you can increase your AdWords conversion rates by 3 o’clock this afternoon.  This is the stuff I experimented with, and I went from a 12% conversion rate to a 56% conversion rate and it now performs better than our organic traffic most of the time. Read the rest of this entry »

10 Comments »Tags: conversion optimization, Google AdWords, increase conversions

Last week, I wrote about how startups should gather all the analytics they can.  This week, I want to give you a couple of free marketing tools you can download to help you tell your story in the next board meeting.  Hey, we’re a startup too and this is helping us, so I figured why not share it.

Free Marketing Tool #1

Read the rest of this entry »

2 Comments »Tags: free marketing tools for startups, startup marketing

We’ve been paying a LOT of attention to the folks who use our product as we build out a minimum viable product, along with a scalable and repeatable business model.  Since July 12th (roughly 6 weeks), nearly 6,000 people from just about every country on the planet has used our SiteJuice product, running over 10,000 websites through our system which shows them basic SEO information with resources on how to fix various issues.  We’ve collected close to 500 responses to our KISS Insights survey asking people what they thought – how useful was the information we gave them.  The answer was overwhelmingly “very useful” – but that doesn’t mean we’re ready to throw up a paywall just yet.  Let me explain…

Read the rest of this entry »

6 Comments »Tags: minimum viable product, startups

You’re a startup and it’s you (the CEO), your CTO and your marketing guy in the monthly board meeting, and your investors ask “so what did you learn from the marketing activities last month?”  Don’t say something like “we’re not real sure”, or “traffic went up, but we don’t know from where” – or anything like that.  In fact, make sure you don’t fall into the old adage “I know half of my marketing is working – I just don’t know which half.”

Read the rest of this entry »

4 Comments »Tags: analytics, chartbeat, Clicky, crazyegg, dave mcclure, google analytics, hiten shah, kiss insights, userfly

You might not expect it, but these entrepreneurs each have something to say about customers – the critical component to any startup.

Steve Blank is a retired serial entrepreneur who teaches entrepreneurship at UC Berkley and Stanford University, and the concept of Customer Development which is outlined in his book, 4 Steps to the Epiphany.

Steve clearly defines a startup as an entity who’s job is to discover a repeatable and scalable business model. Ideally you should find this business model before you start investing a significant amount of funds into the business. Steve cites several examples in his book (he really likes the Webvan story from the 90′s) where startups fail because they never bothered to find out what the customer wanted, why they liked or disliked about the product, how they used the product, and what would make them pay money for it (or not).

Steve believes it is crucial that the founders of the startup “get out of the building” and find the needs of the customers and who they are (customer discovery) and then test your hypothesis against a sample group of these customers (customer validation). If the test fails, it means you have to go back to the drawing board again (the “pivot”). You can’t go on and build your customers, and hence build the business, until you have found something scalable and repeatable. Many startups make the mistake of marketing their product heavily and hiring a small army of sales folks (customer creation and company building) to sell the product before they pass “customer validation” piece, and then scratch their heads as to why no one is willing to pay for the product.

Diagram courtesy of stevblank.com

37 Signals founders Jason Fried and David Heinemeier Hansson, who just released their book Rework, have a different take. They’ve built a business on developing a product that THEY want to use, instead of listening to what everyone else wants.  They talk about how to “say no” to customer requests, instead of being pulled in fifty directions by listening to them.  They talk about “scratching your own itch” and building a product you would use yourself.  “When you solve your own problem, you know exactly what to do.  When you solve someone else’s problem, you’re stabbing in the dark,” says Fried.

The book offers several concepts on how to build and run a small business while defying everything you think is “normal”.  Over 1,000 people a week sign up to use their products, Basecamp, Backpack and HiRise and have done so without any outside investment.

Then you have Mark Cuban, a successful entrepreneur who did well during the dot com boom and is now the current owner of the NBA basketball team Dallas Mavericks.  Usually cited for his brash remarks, Cuban talks about a variety of things on his blog from the stock market to Facebook privacy.  I recently found this on his bog the other day: “Why You Should Never Listen to Your Customers“.  Cuban basically says that entrepreneurs far too often try to emulate their competitors down to the feature, which prevents them from inventing anything truly unique and different.  So now when a customer requests a feature, its usually because of a feature they saw on a competitive product that they wish your product had.  He says “its not the job of your customers to know what they don’t know” and that every entrepreneurs job is to “invent the future.”  Steve Blank makes a similar comparison in a blog post entitled “Death by Competitive Analysis” where he says “instead of optimizing for a minimum feature set (as defined by customers), a competitive analysis drives a maximum feature set.”

These are all different views on how finding and listening to customers can affect your business.  There may not be any one right answer, but in the end, we think you should definitely listen to your customers, and that you (the software developer) should be a customer of the product.  We believe in the concept of eating your own dog food, rather than trying to build a product in a vacuum.  At then end of the day, we think it’s important to have an active customer feedback loop, but be careful not to go off into the weeds building one-off features that will only be useful to a few users.

No Comments »Tags: 37 Signals, customer development, David Heinemeier Hansson, Jason Fried, Mark Cuban, Steve Blank