PadiAct Update – Multiple segments, New segment rules, Custom actions
Last month we have been pretty busy with the new developments on our behavioral targeting platform, PadiAct. As our initial experiment with targeting newsletter subscribers went really well, we have decided to take the platform further so we can play some more. 
Here is what we have been up to:
Multiple Segments Targeting
We are adding the possibility to define multiple segments for the same website. Let’s say that you would like to greet your organic traffic with one message and the paid traffic with another one. Or maybe you want to offer a discount to your returning visitors and invite your new ones to follow your social media profiles. The platform now supports unlimited segment and action definitions. However we have been testing with maximum 3 actions as, after this level, the JavaScript files tends to get above 20Kb.
Here is a small glimpse of the segment management page:
Fully customizable actions
Well, almost. While we are using fully customizable actions in our sandbox environment we still need to come up with documentation on how to use it. At this moment we activate it only on request base.
However we are going to roll out the possibility for deploying any message to a certain segment of your website. We allow HTML so you can also add forms, images, video or just plain text with links.
New Segment Rules
The following segment rules are going to be rolled out:
- New vs Returning Visitors
- Visitors using the search functionality of the website (based on the keyword they search)
- Visitors visiting a certain page of your website
- Visitors landing on a certain page of your website
- Visitors spending a certain time on your your website
An important requested rule by our beta testers was to be able to specify the pages on which the script should not run. The reason behind this rule was to not distract a user once he entered, let’s say, the buying process or the creating an account one.
Another requested development that we have implemented was to allow RegEx expressions when defining segmentation rules.
Of course the new update will come with bug fixes and some usability improvements based on the feedback that we have got so far. It will be rolled out around 1st of May, so stay tuned.
Form tracking and optimization
On the web there are 2 main ways a visitor can interact with your website: through links and forms. Think of it as an asymmetric dialogue. Through links you offer information to the user while through forms the user sends information to you. Being asymmetric is what sucks about it, but than again, if there would have been an easier way, I am sure it would of have been implemented.
Most web analytics tools tend to track the outcomes of a visitor actions: the pages that load after users click a link or fill in a form. While for links I believe that is the right way to go, with forms things stay different. No matter how well optimized, there is always going to be an abandonment rate. With forms, things are a little bit more complex. So, how do we track them?
Tracking form errors
Tracking form filling errors is the first thing you wanna do when you want to know more about the performance of your forms. Make sure you have implemented a form validation method though. Check this tutorial on how to do it if you haven’t done it yet.
Toward Behavioral Intelligence
ReadWriteWeb posted today a great video made by IBM, called “The Internet of Things”. Here is the video:
Back to the basics: search engines & user engagement
There’s been a lot of web ink spilled on the subject of measuring user engagement. It is true there is no magic formula for measuring it, but take my word for it: when a user buys something from your web store, fills up a lead form, subscribes to your RSS feed or comments on your latest blog post, that user engages with your website.
Try to answer the following question: how well is your website doing when it tries to engage visitors that have a clear focus? I am talking about those visitors that land on your website by searching for a specific keyword on search engines, unrelated to your brand.
1. Start by setting goals for each action you deem means engagement
The following video will offer you clear details on how to do exactly that:
If you could ask one question to a certain segment of traffic, what segment would you choose and what question would it be?
When it comes to behavioral targeting, I believe in simple solutions and to simple solutions you get by asking the right questions.
The question: If you could ask one question to a certain segment of traffic, what segment would you choose and what question would it be? The guests: Avinash Kaushik, Joseph Carrabis and Stephane Hamel, people I look way up to when it comes to web analytics brain set. Here is how they would answer the above question:
Avinash Kaushik
Question: Why were you not able to complete your task on our website today?
(Author’s note) I love Avinash’s “simple solution – rich in value” approach. Implementing his suggestion is almost piece of cake to do with some minor help from your development team; the value you can get out of this implementation, priceless
.
Joseph Carrabis
In most cases the visitor percentage goaling/converting is small (just checked FireClick. Conversions are 4.2% and cart abandonment is just over 70%). These numbers are extremely consistent with NextStage research and are due to site designs/topologies/… not being designed for the majority audience (except by accident, usually). These numbers (4.2%, 70%) demonstrate a site’s de facto audience. They’ll complete a goal/convert regardless.
I should point out that I’ve written elsewhere that NextStage has a 19% bounce rate, a just over 75% conversion rate and a cart abandonment rate of 3.7%. The low numbers experienced by other sites don’t surprise me because few sites are designed to capture the majority of their visiting audience.
Essentially these sites don’t ask visitors “Is there anything we could do to help you achieve your goal on the site?” and note that I’m asking the visitors how to achieve their goals while on the site, not how to get them to achieve the site’s goals for visitors.
Stephane Hamel
Part of the RFM segmentation model (Recency, Frequency, Monetary), Recency would be a better predictor of behavior than any other metric. Think of it, just with this metric you can also segment new vs repeat visitors, prospects and long-time customers, unqualified visitors vs advocates, etc. Segmenting by Google search terms are interesting, doing it by social media source is cool, but nothing beats Recency. Philip Kotler, the gold standard in marketing management, sums it up nicely: “Marketing takes a day to learn. Unfortunately it takes a lifetime to master”… Have you checked your recency metric recently?
Now, the question I would ask returning visitors is very simple: What is your purpose of visit. Knowing that specific piece of qualitative info would allow me to see if/how I can help achieve their goal – and mine!
My Take
I love all 3 answers, and to prove it, I am going to implement them all. I’ll let you know the results as soon as I will have some.
I like Avinash’s answer cause it’s so plain simple to do and results can be very helpful. I love Joseph’s answer cause I don’t think you can get a better way to learn on how to optimize a conversion process; it really can’t get more relevant than that. I love Stephane’s answer because when visitors return over and over again to your website you can bet they have a good reason for it. You’d better find it out.
Your take now
If you could ask one question to a certain segment of traffic, what segment would you choose and what question would it be?
Give us your best shots and we’ll choose the ones we can accomplish and offer their implementations for your website, free of charge.