How to Squeeze Every Drop of Valuable Visitor Data from Your Forms (Without Complicated Programming)

It’s the great marketing catch-22: You need visitor data in order to create value-filled, personalized, highly targeted offers. The more information they provide, the greater your ability to build personas that fit your ideal customers.

But your visitors still aren’t convinced. Like jittery fish, you’ve presented them with the tastiest bait, and they’re hesitant to even take a nibble. So how can you get the information you need without scaring them off?

The answer is…

Progressive profiling.

What is Progressive Profiling?

Progressive profiling uses dynamic form fields to ask for and collect information on prospects based on the information you already have about them. Like a first date, it gets to know new customers in a way that’s gentle and unassuming. What’s your name? What’s the best email address to contact you at? How can I help?

Then, as the customer’s interaction with your product or service continues and greater trust and brand recognition is built, more questions are asked – helping you not just capture leads, but build on the intelligence you’ve gained.

Progressive profiling may start with name, email and question, but can gradually lead to vital details that will help you better understand where your prospect is in the buying cycle: are they evaluating products? Comparing features? Focused on pricing?

Asking questions about their timeframe to purchase and where they are in the decision-making process can help you better understand how and where your product fits according to their needs. Understanding how they plan to use the product can give you valuable insights into tailoring your offer across every stage, leveraging the information you already have about them.

How Does It Work?

Let’s say you sell a product and you have an email newsletter filled with valuable tips and techniques on using it. As is often the case, your prospective customer is asked for their name and email address. Simple enough, right?

They subscribe and get their first couple of newsletters. They’re starting to feel comfortable with your brand and enjoy what they’re getting so far. Here’s where the progressive part comes in. You’ve got another freebie for them – you just need a bit more information about how they plan on using your product, when they plan to purchase, and approximately how much they plan to spend.

Since you don’t need their name and email address again, the form simply asks the relevant questions you’ve set up. This information helps you gradually begin to understand this customer based on how much they’re willing to share as your relationship with them progresses.

Obviously you’ll want to space these out and give as much as you get – take the time to answer questions, clarify features and options, and so on. Don’t expect the customer to be receptive to answering personal questions by day 2.

Are you interested in applying progressive profiling? Kissmetrics is a great tool to help you with progressive profiling. We collect customer and behavioral data that you can use to optimize your forms. Try Kissmetrics today!

Don’t Forget Your Best Practices

In addition, you’re not going to want to throw best form practices out the window when implementing progressive profiling. Things like:

Being Clear about What Information is Needed (And How It Should Be Presented)

This example from Threadless tells customers precisely how to enter their payment information.

Explaining, if Necessary, Why Certain Information is Needed

The example below, from Money Supermarket tells people why they need the registration number of their car.

Source: eConsultancy

Make sure your reasoning is reasonable – Money Supermarket could have said, “We need your information to help get you an accurate quote”, but it still wouldn’t answer the prospect’s why? The fact that they need it to find that same exact car and get you a quote is much more sensible.

What Happens After They Subscribe?

And finally, don’t forget to be clear about what happens after the sign up process is complete. How can they download the freebie? How often will they receive email newsletters? Can they unsubscribe or change the frequency? And so on.

BufferApp sets newsletter expectations before asking for subscribers’ info

How Can You Implement Progressive Profiling?

Progressive profiling is a relatively new function, but there are a variety of solutions on the market that make integrating it into forms easy and hassle-free. Oftentimes, you can do this without any programming knowledge at all. Here are a few services that offer Progressive Profiling as part of their systems:

Salesforce Pardot

Adds a progressive profiling feature that ties in with your existing use of Salesforce Pardot

HubSpot

HubSpot refers to progressive profiling form capabilities as “Smart Forms” and provides a variety of best practices you can follow when implementing them.

Act-On

As with other services, Act-On lets you define a set of rules for which other form fields will display. The site also provides an example tour through how its Progressive Profiling system works.

JumpLead

An example of the JumpLead dashboard

JumpLead is a sort of marketing automation/lead generation platform of which progressive profiling forms play a role. Like many of these other solutions, they offer a full scale of marketing automation tools. For WordPress users, JumpLead also has a plugin that can integrate progressive profiling into their existing WordPress system.

How to Get the Best Possible Results from Your New Progressive Forms

It’s easy to fall into the tempting trap of asking for more and more information through more and more form fields as your relationship with the prospect progresses. However, small bits over time (and depending on the customer’s stage in the product’s overall lifecycle, as well as their buying cycle) will help foster a reciprocal relationship while giving them the personal attention and nurturing they crave.

And don’t forget, progressive profiling is just one tool of many designed to help improve your forms and increase your conversion rates. As with any tool, it’s not a silver bullet – it’s all in how you use it. By making progressive profiling a part of your overall sales and conversion optimization process rather than looking at it as “yet another form component”, you’ll be poised to start forming lucrative customer relationships built on a foundation of mutual trust, understanding and expectations.

Now It’s Your Turn

Are you using progressive profiling in your own lead generation forms? How has it worked out for you so far? Share your success stories and triumphs with us, as well as your thoughts on this unique practice in the comments below!

About the Author: Sherice Jacob helps business owners improve website design and increase conversion rates through compelling copywriting, user-friendly design and smart analytics analysis. Learn more at iElectrify.com and download your free web copy tune-up and conversion checklist today!

Board Strengthens Appraiser Qualifications

The Appraiser Qualifications Board (AQB) recently updated the minimum national requirements for becoming a real property appraiser in an effort to continue to advance and attract the best and brightest to the profession.

Fox News releases Bob Beckel over his 'personal issues'

Bob Beckel, a Fox News host who has struggled with prescription drug addiction, was let go Thursday in a move that a top executive blamed on Beckel's "personal issues."

Getting Your First 1,000 Facebook Fans (Infographic)

More and more people are using Facebook to research businesses these days, and they’re doing more than just finding a business’s location and reading posts. People are using Facebook for social proof. They want to see how many people “Like” a business.

Think about it. If you want to go to a new restaurant and you visit its Facebook page and it has 30 fans, does that make you more or less likely to visit the establishment? It probably doesn’t help. The same goes for internet businesses. If you’re running an ecommerce store and you have only 100 fans, do you think that helps your chances of getting orders, especially when your competitor has 5,000 fans?

This doesn’t apply to all businesses, but for most B2C companies, having a lot of Facebook fans is important. This is especially true for small businesses. In certain bigger industries, such as automotive, the amount of Facebook fans doesn’t factor into a buying decision for the consumer.

To help you grow your fan base, Neil Patel over at Quick Sprout has created an infographic that’s full of wonderful tips to help you reach your first 1,000 Facebook fans.

Courtesy of: Quick Sprout

For additional tips on getting more Facebook fans, check out this content from Kissmetrics:

• How to Get More Likes on Facebook (Infographic)
• How to Increase Facebook Fans and Twitter Followers

About the Author: Zach Bulygo (Twitter) is a Content Writer for Kissmetrics.

Gen-Xers don't want to be homeowners

Millennials aren't the only ones hurting America's homeownership rate.

Australia is selling a ranch the size of New Jersey

Read full story for latest details.

Ben Affleck controversy causes PBS to delay 'Finding Your Roots'

PBS announced on Wednesday that it will postpone the third season of "Finding Your Roots" following an internal review that concluded actor Ben Affleck improperly influenced the show to omit the fact that his ancestors owned slaves.

Using the “So What?” Test to Transform Your Copy

Take a look at various products and services around the web, and you’ll find one common theme.

They all talk about themselves and what their product does. Few of them talk about what they can do for you. This is a devastating mistake that marketers and copywriters make.

To write effectively, you can apply a simple test to all your copy.

The “So What?” Test

You want customers, right?

Then start telling them what your product does for them!

Stop simply reciting what your product does. For every line of copy you write, ask yourself “So what? What does this do for them?” Let’s use Dropbox as an example:

Dropbox stores your files in the cloud.

So what?

It means you can use Dropbox and retrieve your files from any device. The files don’t have to be stored on the device.

Now put that in a short sentence.

Dropbox keeps your files safe, synced, and easy to share.

Okay, great. Now I understand what it does for me.

What Does Your Product Do for People?

Customers don’t care about your product itself or how hard you work on it. They care about what it does for them.

So what does your product do for customers? What are the benefits, and how do those benefits solve problems?

Let’s take a look at a couple of products.

Salesforce

Check out this copy for Salesforce:

It’s a simple, benefits-focused opening sentence. They don’t write about the product. They write about what it does for people: it helps them sell smarter and faster with the biggest CRM in the world.

They could write this:

Salesforce puts all your leads and sales into one database.

So what?

What does that do for me? I can just use a spreadsheet for that.

Let’s move further on down the page, and you’ll see that it’s all about the prospective customer and what Sales Cloud can do for them:

Sales Cloud handles a lot, so you can take on even more.

The [what the product does], so you can [how this helps you] is a good template for copy. Explain what your product does, and then explain how it helps the customer.

As you can see above, throughout the rest of the page, all the copy is benefits-driven. Even the video isn’t about the product. It supports the top line copy (sell smarter and faster) with a testimonial from a customer.

Let’s take a look at another product, this one in the B2C space.

Sonicare Toothbrush

Toothbrushes are a big market. The majority of people on Earth need one, and they will need to keep repurchasing them until they die. There are a lot of toothbrushes out there from many different companies.

So if you’re responsible for marketing a toothbrush, how do you make yours stand out? Philips Sonicare has done a good job of this. Let’s examine the copy:

People don’t care about a toothbrush. They care about what it can do for them, the effect it can have on not just their teeth, but their overall health and appearance.

In the above three sections of copy, we get concrete numbers on why the Sonicare toothbrush is better than the traditional manual toothbrush. It removes 7x more plaque, improves gum health in 2 weeks, and whitens teeth 2x better than a manual. This is what people care about – healthy, shining teeth. They don’t care about how many bristles are in each brush or how the toothbrush works. They care about what it can do for them.

The remainder of the page discusses some extra features of the toothbrush:

Do you think Philips hired a copywriter to write this? Probably not. There are no clever headlines or fancy writing here. It was likely written by someone on the product marketing team. Does it matter? No.

You don’t need to hire a copywriter in order to put together an effective landing page. Just speak directly and tell the reader what your product can do for them. You’ll be putting yourself in a good position to boost conversions and bring high quality leads to your sales team.

Hat tip to James Currier for providing us with the “So What?” framework.

About the Author: Zach Bulygo (Twitter) is a Content Writer for Kissmetrics.

The middle class is struggling to make the rent

The combination of strong rental demand and low vacancy rates has made affording the rent harder.

Boeing names new CEO

Boeing is bringing in fresh leadership -- the company named a new CEO Tuesday.

How to Take Advantage of Machine Learning and Kissmetrics to Reduce Churn

Oh you, young and motivated SaaS company. I know you very well. You have a disruptive way of seeing things, a cool product, reasonable prices (well…), and just want to make the world a better place. But you won’t be able to do that if you don’t try to understand the behavior of your hard-earned customers.

You know (well informed that you are) that the “Leaky Bucket theory” is a real thing. Acquiring a new customer will cost you more than trying to keep the ones you already have. And that’s where Kissmetrics in-app analytics and machine learning can help you make better-informed decisions for your customer retention campaigns.

If you’re reading this, I must assume that you already have a fairly good idea of how Kissmetrics works, but, you may have encountered some problems when trying to set machine learning on top of it, and get precise answers. And that’s what we are going to be talking about today: “For machine learning to answer my burning questions and help me reduce customer churn, what should I track in Kissmetrics?”

The good news is that it is easy to get your Kissmetrics tracking plan well suited for machine learning (in case you weren’t following earlier, customer retention goes hand in hand now with predictive analytics and thus, machine learning). With just a few tweaks to your Events, Funnels and Marketing Campaigns, you’ll be getting the exact answers you’re looking for to keep your customers happy and drive growth.

So…follow the lead!

What you need to do first: Prepare your dataset

First thing’s first, you need to know that Kissmetrics lets you export your data. If you think you can manage this on your own (for this part of the process), then just skip this part and jump straight to: “Enter the fray: Define your targets and events”.

For all the others that stayed with us, let’s press on. So, Kissmetrics has a Data Export feature that’s easy to set up. If you’re already lost, check this link and then come back to this section. Interesting enough, your data even goes to S3 on Amazon.

Anyways! After exporting your data, you will get files with JSON lines. Amongst these lines, some are events, some are properties settings, some are aliases…etc. There is definitely some data preparation work to be done here. In particular, you need to reconstruct aliases history, and gather your users’ properties.

It might not seem very user friendly at first, but this export format has a crucial advantage for machine learning. Indeed, most analytics apps map user properties for a single day. But machine learning needs to access user data over a large period of time. The longer the available period, the more accurate the predictions on the user’s behavior will be. And this is just what I love about Kissmetrics compared to other analytics: you can reconstruct the properties state of every user at any point of time!

Once you’ve built nice data tables, you need to define how far you will look into the past, how far you will look into the future (who will churn in the next 7, 14 or 30 days?), and you need to gather your targets (aka the answers to the above questions). Only then will you be ready to use machine learning. So, without further ado, let’s move on to defining your targets and events.

Enter the fray: Define your targets and events

Machine learning is good at answering binary questions: “Will my user convert or not”, “Remain active or not”, “Upgrade or not?”, etc. Ask the machine and it outputs a ‘score’ for every single customer. A score is a probability for the answer to be “yes”. The higher the score, the more likely your user is to behave like the target (got it ? :P).

Step 1: List out what you need to predict

The first rule to uphold is: track anything that describes the target’s behavior, i.e. what you want to predict. To predict conversion, you need to define a conversion event, and to predict churn, you need an unsubscription event. Make sure these events are included in your tracking plan.

For all ‘retention’ type questions, you’ll need to choose a global activity metric based on your events. Classically, you can use stickiness, number of events, engagement indices or power usage.

If you choose stickiness, you ideally want to track daily visits to your app. “Log-in” events are tricky since a user can be logged in for several days without logging out. So instead, some companies track massive events with very little information, like “Page View”. But “Page View” is expensive for most tracking plans and too general to help. My advice is always to spend your money on valuable events, such as your key features and funnels! You’ll miss the few visits where users did not trigger any of your interesting features, BUT you will focus the machine on what really matters in your app.

That being said, be careful not to fall into the trap of overly specifying events that are too rare to build reliable statistics on, or too numerous to be understood by a human – or sometimes even by the machine.

Step 2: List your app’s key features

Machine learning is not magic and it’s not a palm reader either (even if we want it to be just that). It predicts the future by analyzing the past, and looking at usage patterns which drove certain behaviors or targets (e.g. conversions, upgrades, etc…). Then, based on recent behavior, it computes the chances of observing the target in the future. The word “behavior” is important. The machine reads behavior in your events, and the success of your predictions clings to your tracking plan and its precision.

Obviously, business outcomes such as conversion, upgrades, or churn will depend on how users interact with your app. List the key features in your app and make sure that each of them is properly tracked with one specific event.

For better readability, I strongly suggest naming events by clearly referring to their respective app features. For instance, don’t name a click event by its button shape (e.g. “right corner red button”) but rather by its functionality (e.g. “delete project”). This way the machine could output “Users which delete projects more than15 times in the last 3 days are 2 times more likely to churn”. Good naming will also help you in your daily usage of Kissmetrics.

But don’t be too hasty when creating your tracking plan! Not all events are valuable. It seems reasonable to say that “You know that an event is valuable when, by removing it, you lose global value”. Losing value in machine learning means degrading the predictions. So a valuable event is an event that drives predictive power. Yep! Sometimes, more is actually less. ;)

So in short:

• Track anything that describes what you want to predict
• Track all the key features of your app
• Don’t use too generic events
• And don’t be overly specific in your events’ definition

If you want more information about event tracking, Diana Smith from Segment gave a great presentation a few weeks ago. You can find it here.

All of this constitutes your tracking plan, so you can now move on defining your funnels.

Step 3: Define behavioral funnels

I am sure you have defined marketing funnels in Kissmetrics, such as a signup conversion funnel “Viewed HomePage -> Viewed Sign Up Page -> Completed Sign Up”. Marketing funnels are key to monitoring your main business KPIs.

From a product perspective, behavioral funnels are valuable to measure task/feature completion in your application, e.g. Added Member -> Created Task -> Assigned Task in a project management app. Analyzing these funnels helps you point out, on your user paths, actions driving engagement and those causing frustration (eventually churn).

Behavioral funnels will require you to track more events, and can turn out to be expensive. However, I recommend that you keep them in place as long as they are notably improving your predictions.

Step 4, AKA The end of the road: Make your tracking stable

As stated before, machine learning builds models based on past events and looks at present events to predict future events. Changes in your event tracking plan may harm your predictions. Therefore, it’s very important to carefully plan the redefinition of each event. Ideally, events should only be changed when major versions of your service are released. I would recommend to:

• Introduce versions of the same event “feature1_v1″, “feature1_v2″…
• Spot the events that are bounded to be unstable and suffix them with “_noML”. Thus ignoring them in the machine
• Ignore the adding, removing or renaming of an event (new or old feature) for some time, or ignore it retroactively
• Lower the level of detail of your targets in unstable times, e.g. “Pay Event” instead of “Pays $19 Event” will make transitions smoother between Pricing Plans.

After all these magnificent tips, you’re ready to use Machine learning. There are Open Source Libraries (Scikit-learn, Shogun, Mahout, Spark MLlib…), and Predictive APIs (PredicSis API*, Google Prediction API, prediction.io…) here to help you out. Several SaaS companies have recently appeared to offer an end-to-end service, from Data Export to Scores (ChurnSpotter.io*, Preact.com, Frontleaf.com…). Keep an eye out for them; some of them already support Kissmetrics.

Enjoy the fruits of your labor: Track your campaigns

Finally after this long journey of events and tracking, you hold your scores in hand, AKA the answers to questions like “Who will stop using my app tomorrow and why?”, or “How many return customers can I expect?”. One score per user depending, of course, on the chosen target.

Anyway, now is the time to take action on your scores and to convince the hesitating users to become great customers.

First, push the scores to Kissmetrics as user properties so that you can segment users by scores. My advice is to round scores in order to get 10 segments “0-0.1″, “0.1-0.2″, …,”0.9-1″.

Suppose you need to increase your expansion sales, and you’re sending emails for an upgrade campaign. You will use the “upgrade” scores, which reflect the propensity of each of your users to upgrade. In Kissmetrics, track your campaign and build your campaign report by segmenting by scores, and filtering on people to whom you send the email. Build a control report, also segmented by scores, this time, filtering on users who did not get the email.

You will probably notice that very high scores upgrade anyway (“loyal” users), and very low scores do not upgrade whether they were sent the email or not (“lost” users). However, your campaign was effective on middle scores (“undecided” users), and you learned (wait for it…) who are the “undecided-but took-the-right-decision” users who compose your target for this particular campaign.

The next time you send emails, you can differentiate your message based on scores, or adapt to the likelihood of your user to upgrade. This allows for a better communication with each user, the detection of dissatisfied users early enough, and the reduction of the marketing pressure thanks to optimized targeting. That is how machine learning will boost your upgrades hence your expansion MRR!

Conclusion

As the saying goes: “A picture is worth a thousand words”, so, without further ado, here is a little picture summarizing the idea behind this article.

Machine learning has never been so accessible to non-machine learners; and could bring significant benefits to your business once fed by analytics. You’ve already gotten off on the right foot by using Kissmetrics, now get the machine to help you anticipate customers’ behavior, and move to data-driven predictive marketing!

* Full disclosure, I work for PredicSis as a machine learner, and participated in the birth of our ChurnSpotter product, which do all the hard work for you and enable you to better retain your users.

About the Author: Florence Bénézit currently works at PredicSis on the ChurnSpotter.io project. She holds a Ph.D. from EPFL, Switzerland, in Distributed Signal Processing. She has been working as a Data Scientist in the industry for the last 4 years.