Content consumption does not always equal purchase intent

Content consumption does not always equal purchase intent

Modern marketers can all agree that content marketing is the key to engaging and acquiring new leads.  A great thought leadership webinar can attract new prospects to your solution and educates them on the benefits it can deliver.  The typical wisdom of marketing automations systems is to award points to leads  for consuming the great content marketers publish.

I’ve implemented many lead scoring models in my time.  I’ve built out smart campaigns to award points for white papers downloads, points for attending a webinar, more points for visiting a website.  More points get awarded for the right job title and maybe even company size.  The leads with the most points are dubbed “qualified” and make it into my sales teams’ queue.  

Over time the sales team learns to pick and choose which leads they want to call based on what they have learned from previous calls and which leads to pass over.  They know the lead is just downloading content for the sake of education and is not ready to talk to sales.  I’ve seen sales teams gradually become less and less inclined to look at the lead score and just “go with their gut”.  And at the end of every quarter the head of sales inevitable asks me:

“Is our scoring model right?”  

“Why is our qualified point threshold at 50 points?”

“How do we know a white paper is worth 10 points, or 20 points, or -17 points?”  

The answer is that without some heavy statistical analysis, and regression testing, a lead scoring model is really just based on assumptions we marketers make.  They may be highly educated assumptions, but assumptions none the less.  A good lead scoring model needs to take into account more than just activity on the site and consumption of content.  Downloading white papers and attending webinars certainly works to cultivate leads and educate them about why they need to purchase your product, but be careful to not confuse correlation with causation.  They may score all the behavioral points they want but will never be the right fit to actually purchase your product.

An even better lead scoring model should be predictive.  Meaning, a predictive model is an analysis of all closed won deals to find commonalities amongst those leads.  Once the model knows which is the ideal lead, you can prioritize those leads and optimize your sales team’s time.  

If you want to make your sales team happy, send sales leads that match contacts that have already closed.  Not just leads that downloaded a bunch of content.  

4 Account Based Marketing list building tips

I’ve given 20+ presentations on how to implement ABM and have shown a version of the slide below every single time. In the most overly simplified approach there are 7 steps to building and running account-based marketing.

And when I say overly simplified, I really mean it. The first step is to identify your account list. This step in implementing ABM can take months, even the better part of a year.

Even here at RollWorks, where ABM is at the core of our business, this process has taken over 4 months. In the course of those months I learned some things that I honestly wish someone had told me earlier. So here are my four learnings from picking our target account list at RollWorks.

Lesson #1 Decide as a team why you are picking a target account list

I initially started having meetings with our Head of North American Sales in September of 2017 to discuss building a target account list for our B2B sales team.

That first meeting quickly snowballed to a larger session with our Head of Revenue Operations. Then we added a Business Intelligence Analyst, Sales Compensation Analyst, two sales leaders, and my boss, the Director of Marketing.

I quickly realized that my desire for a target account list spanned across our organization, which was fantastic in theory, but soon I found out that everyone had a different notion of what the account list would be and the purpose it would serve.

The Sales Compensation Analyst wanted the list for building a forecast.

The sales leaders wanted the list to fill up their reps “book of business.”

I wanted a target account list and a scoring model to rank inbound inquiries so I could serve up the best leads to sales.

So my first tip to picking a target account list is to establish a clear objective and end goal for the project. That way all members of the project will have a clear understanding and expectation for how the list will be used as well as the outcomes of the list.

Will it just be a list? Or will there be a scoring model that helps to identify net new accounts that match your ideal accounts over time?

Lesson # 2 Define your ABM Terms

We held weekly sales and marketing alignment meetings to pick our target account list and build our plays. In the course of those weekly meetings the team discovered that we all used different nomenclature to describe what we were talking about.

Often times in meetings we found ourselves spending 5-10 minutes just explaining the differences in terms like “total addressable market,” “owned accounts,” and even “named accounts.”

To an outsider, things like “priority, tier, and rank” may seem interchangeable, but for our team we wanted to get clear on the terms so we could all understand what each other meant. Coming up with our common definitions was honestly integral to keeping our sales and marketing meetings moving forward, so we devised some definitions and even created graphics to pass around the team to keep everyone on the same page.

TAM = Total Addressable Market

This represents the size of potential customers of your solution. This can include companies that use your competitor, use a certain technology, companies of a certain size, etc.

ICP = Ideal customer profile

This should be the firmographic, technographic, and business signals that make up the best customers for your product.

NAL = Named account list

The top accounts that match your ICP, that are in your TAM, and have a high likelihood to close.

Tier = Ranking of the company

This can be based on a number of internal factors such as employee size and media spend potential.

 We assigned 5 tiers to our accounts to differentiate between enterprise, mid-market, emerging, and SMB sized accounts.

As with defining the goals of why you are building a target account list, I recommend making your own dictionary of frequently used terms so that everyone on the team knows the difference with all the acronyms.

Lesson #3 The past, the present, and the future of a target account list

Another interesting development of building our target account list was that by the time we got it into our reps hands, things had already shifted. For example, we hired four additional reps and didn’t have enough accounts to fill up their territory. Another issue was that as we were building the list, some target accounts had already turned into customers.

What I realized is that we had failed to think about the future state of our list. Some useful questions to ask as you go through this process:

  1. What is our process for adding new accounts to this list?
  2. How do we remove accounts from this list?
  3. If we close one account, do we add more or just keep going with the same list?

Answering these questions will help set up the broader team with an ABM strategy that works for the full year, not just one quarter. In our case, our team decided to refresh the list quarterly with additional accounts as our sales team worked through their lists.

Lesson #4 Don’t forget about the people

The last stumbling block we encountered was that we frankly forgot about the people. Yes, the strategy is called account-based marketing, but we still communicate and work with people within the account.

Since a good portion of our time was spent figuring out the signals and data points of our target accounts, we honestly ran out of time to make sure we were finding the right people to target as well.

Luckily the RollWorks platform is cookie-based so it can find and target the people within the accounts after we’ve chosen them. If you’re using other (IP-based) ABM providers this might be something you need to consider. We ended up using our lead locator product to target key contacts at the companies on our target account list. 

My recommendation is to figure out your key personas within your target accounts and identify at least three key personas that have an influence on who buys your product. You’ll want to work through what are the key needs, pains, conflicts, goals, and desires of each persona. This is honestly some good old fashioned marketing work and it really helps to outline the content needed to educate and move those key personas through the buying decision.

How to evaluate a predictive marketing vendor

Predictive Marketing Vendors

After working for two predictive marketing vendors, I feel I am in a unique position of understanding a thing or two on how best to pick a predictive marketing vendor. If you’re new to the space, it can be difficult to pinpoint key functionality that will align with your company’s sales and marketing objectives. I’ve compiled quick list of the Dos and Donts of evaluating a predictive marketing vendor.

#1. Do establish your company’s needs and initiatives for predictive marketing
Scope out what you are trying to achieve internally first. Is it improving your sales win rate? Improving lead conversion rates and velocity? I’ve identified 6 primary use cases for predictive marketing: sales productivity, marketing campaign effectiveness, sales forecasting accuracy, demand generation, database segmentation, and cross sell / product penetration.

#2. Do evaluate only the vendors that address your company initiatives
Some predictive vendors excel at helping you identify patterns in your anonymous traffic, others at sourcing raw leads for your database. Pick a vendor that aligns with your company initiatives.

#3. Do perform a live test of the scores
The purpose of a predictive marketing vendor is to predict actual events, not just how well they can predict what you have already done ( closed, meetings booked, conversions ). A vendor can easily built reports to show how well they would have scored leads and opportunities in the past  leveraging historical data but it is entirely a different thing to score leads in a live application.

#4. Do pick a vendor that has the expertise and best practices to train and onboard your team for success
As this is a fairly new technology, you’ll want hands on guidance on how to integrate the scores in your workflow, train sales reps, and best practices documentation. Look for a vendor that has published guides, webinars, and success stories

#5. Do pick a vendor that supports integration with CRM and marketing automation platforms
You’ve probably spent a lot of time building out your CRM and Marketing Automation systems.  Best to pick a predictive vendor that plays well with the systems you already know and use.  No need to reinvent the wheel here. Ideally they should have an easy API integration into your CRM and MAP and not require an extra seat nor take up too much at a storage.

And now for the don’ts

#1. Don’t pick a vendor based on the back test alone
Make sure the vendor can perform in a real world application. A vendor that won’t do a live test of their scores is one that probably won’t perform well in real world application.

#2. Don’t rely on a CSV file to build models
CSV files flatten the data and bring forward information in time. When you give the predictive model information that it would normal learn at a later stage in your lead lifecycle, the model wont be able to accurately predict leads earlier in the lifecycle. The best analogy I can give is it’s like Biff in Back to the Future betting on who is going to win the football game with the sports almanac from the future.  In short CSV file will give the predictive model information it shouldn’t have at certain time periods for predicting conversions.

#3. Don’t compare two conversion rates from two different sized sample groups
I call this an apples to oranges comparison.  One sample can have a 36% conversion rate and the other can have a 4% conversion rate, but without knowing the size of the sample group or the denominator to be exact, comparing the two conversion rates is truly meaningless. Remember back in algebra when we were taught to find the lowest common denominator? Same thing applies here.

#4. Don’t make a decision on too little data
Often times I’ve seen prospects attempt to make an informed decisions when only 3 leads have converted to opportunity.  If your average deal length is 6 months and the trial period is only 30 days, you’ll need to pick an earlier KPI in your sales funnel to evaluate performance and make a decisions.  Number of booked meetings might be a good KPI to use.

#5. Don’t compare results from two different time periods
There can be other variables at play such as big tradeshows, higher performing sales reps, seasonality that can impact conversion rates.  If you are comparing the performance of the predictive model against the week immediately following Dreamforce, you may not have as many meetings booked.  Try to compare a similar time frame.  The highest impact tests I’ve seen are when the marketing team splits the SDR team in two, assigns the predictively scored leads to one half of the team and the non predictively scored leads to the other half of the team.  After two weeks they ran the same test again but without the predictive scored leads to make it a true blind test. Then with those results the marketing team could prove to management and sales that calling the to predictive leads first had value.

Building Marketo Bucket Programs for Tracking Content Across Channels

What I have learned over the years is that it becomes really hard and un-scalable to make a new Facebook, Twitter, LinkedIn, or Google AdWords Program in Marketo for every instance of an advertisement of your content. Take for example EverString’s State of Predictive Marketing.  I have that running in retargeting, search engine marketing, Facebook, Twitter, LinkedIn, Nurturing, email list rentals, etc etc. That can quickly get out of control with the number of Programs you would need to build.  Multiple that number against the number of new eBooks and reports and the volume of programs to build and keep track of quickly gets out of hand.

Instead I have built “bucket” programs for both the advertising channel and the piece of content. So for example the State of Predictive Marketing Report, if you download that report through an ad from Facebook you will belong to both the Content program AND the Facebook Master Program. To enable this method I created a Campaign Type in SFDC called “Content” and a Channel tag in Marketo called “Content”.  Then whenever we publish a new report, ebook, case study, video, cheatsheet, etc, a program and corresponding SFDC campaign is built to track all the visits and downloads of that content.

Here is a screenshot of our folder for all our content pieces.  All downloads across all channels are recorded in these programs. We can then easily run reports on total number downloads of a piece of content, pipeline sourced, pipeline influenced by the content.  Previously, I had the content spread across 20+ different campaigns and I had a really hard time calculating how many times a piece had been viewed.

 marketo bucket programs

Then here is a screenshot of on of our the advertising bucket programs.  The smart campaign to belong to this program listens for the URL of the landing page to contain Facebook in the query string.  I just know that when I’m creating an ad on Facebook, the URL of the landing page needs to contain the UTM of “Facebook” and this smart campaign will react.

Marketo UTM Parameters

 

Here is a screenshot of my Google Adwords program.  I clone these every quarter so that I can enter in new period costs based on how much we spent in that channel.

 Google Adwords in Marketo

The UTM fields can get replaced time over time as a lead interacts with more content. The smart campaign inside the Marketo program listens for the different values in the UTM_campaign= xyz to give each piece of content and campaign credit.

So for example if the landing page is posted to facebook the value for UTM_campaign=Facebook. But if the landing page is posted to an external email campaign in MarketingProfs, the value will be “12-2-marketingprofs”.  Examples of URLs I have in use:

http://pages.everstring.com/what-is-predictive-marketing.html?utm_campaign=12-2-marketingprofs

http://pages.everstring.com/what-is-predictive-marketing.html?utm_campaign=Facebook

The end result is that as a team we can track how content performs with multi-touch attribution and see if the content’s impact at different stages of our funnel.

How are you tracking content success across your channels and programs?

3 Reasons Why Sales Doesn’t Call Your Leads

There is a lot of talk about Sales and Marketing alignment out there.  I’ll let the other more official blogs sugar coat the topic and discuss joint meetings, defining stages, and teamwork.  After viewing many Marketo instances and working with a number of teams, I’ve compiled a quick list of the real reasons why Sales doesn’t call the leads Marketing sends over.

Reason #1 You serve up bad leads.

No really.  They are bad leads and you are wasting your Sales team’s time.  One of the joys of inbound marketing is that we are attracting prospects to our sites to download the content we painstakingly produced.  The negative is we also attract junky leads. Thankfully marketing automation makes it easy to control the flow of junk leads into your sales team’s hands.  Thing is you need to actually take some time and build in flows to block junk leads.

Data hygiene and data quality should be as integral to your marketing as good copy writing. Take pride in the quality of leads that you send over.

None of these are real leads. They have no place in your MA or CRM.

Junky leads in Marketo

More over you can’t do a lot of the fun things in Marketo with bad data. Your emails wont get delivered. Building segments and personas for dynamic content gets really hard. Lead routing by geographic territory is impossible if you don’t have up to date location data.

Recommendation: Look into adding a honeypot on your site to prevent spam bots from filling out your forms.  Also build in some data management smart campaigns to routinely remove leads with email addresses like test@test.com or FurryKidder@mailinator.com. Look into technology to help augment your lead profile with valuable information such as location, website, revenue, phone number.  There are many companies out there that offer easy to install webhooks for data append.

Reason #2 You’ve never shown sales where to find the “good leads”

Are you actually producing MQLs? Does the form on the website go anywhere? Is there an email alert built to let sales know when someone starts a free trial?

These are the basics.  But I have seen FAR too many instances of Marketo where the forms on the website go no where and Sales doesn’t know when someone takes action.

Recommendation: Make your Sales team lead views in their CRM for “My MQLs” and let them know that is where they can always find their qualified leads.

Lead Views for Inbound Leads

Reason #3  You haven’t given your sales team a process on how to manage leads

It needs to be a closed loop system. They have to be able to reject or accept the leads you send over. Make it simple. Everything should flow around the field Lead Status.  Teach them to use Lead Status. You give them “lead status = MQL” they either accept it and move it to SAL/SQL or reject the lead as Disqualified or Recycle it to send it back into Nurturing.  

Lead Status and Disqualification Reason

How I got to 0 Duplicate Leads #humblebrag

It wasn’t easy. Believe me.  And honestly I started with less duplicates than any other company I’ve ever worked for.

Coming into Fliptop and getting to basically start the Marketo instance from scratch, I knew that I wanted to build things my way (the right way) and that included getting the database as clean as possible.  There are many reasons why a clean, dupe free database is a best practice.

Now there is no such thing as a duplicate free database. Actually, I just thought of what that database would look like. An EMPTY database would be a duplicate free database.

There are tolerable levels for number of duplicates. Having anywhere between 5% to 10% of the entire database be duplicates is tolerable. I started my process with 4,200 duplicates out of an 85k database, representing 7% duplicates.

My first step was the narrow down how leads entered the system.  Do this step first as it is pointless to clean up the database without stopping how dirty data enters the system.  For me this meant switching out all the forms on the Fliptop website and the blog from a Salesforce web-to-lead to Marketo forms.  Marketo automatically de dups leads if the email addresses match. Salesforce does not. I had a bit of a problem with my engineering and customer success teams entering test leads into the system to test our own predictive scoring. I first cleaned up all those test leads and then built a data management campaign to go through on a monthly basis to delete test leads. 

testingscreenshot

Next I gave my sales reps a tool to add leads in with full contact information. I turned to InsideView as they integrate nicely with both Marketo and Salesforce. If the lead already exists in our system, InsideView will update it rather than creating a net new lead. My reps can research leads and add them into our CRM easily without creating a duplicate mess.

After closing down the avenues of how leads go into the system I could next turn to actually de duping the database.

The process to get to a place of zen and zero dups relied in large part on a tool I found a LONG time ago called DemandTools by CRM Fusion.

It is by no means the prettiest tool around but it gets the job done.  The tool comes with pre built “scenarios” you can run to do sweeps of the database.  Scenarios are basically like matching criteria on the leads, first sweep is to find leads with the exact same email address.  Then the next sweep finds leads with the same name and company name.  Each sweep the criteria loosens up, like the teeth on a comb and the matches on duplicates will become less precise. You can also de dupe leads against contacts and even accounts with similar pre built scenarios.

CRM Fusion Demand Tools

I probably ran 10 or more sweeps using their different pre-built scenarios on just Leads and then moved to de duping Leads against Contacts to whittle the list down. In then end there were probably 200+ leads left in my “Possible Duplicates” smart list inside of Marketo that I de-duped by hand. I know this sounds tedious, but when there were only 200 left I felt I could see the light of the end of the tunnel so I went for it.  The result is when I run the “Possible Duplicates” smart list in Marketo I see “No leads were found.”

No duplicate leads

So how do you handle how leads enter your system and managing duplicates? 

My MQL Hack

As the director of marketing at Fliptop I had the pleasure of being not only the primary user of predictive lead scoring for my demand gen efforts, but I also get to work closely with our engineering team to help shape the product. A couple weeks ago we brought on a new lead engineer in the San Francisco office and I gave him my normal “this is B2B marketing” presentation. He asked to see inside Marketo and Salesforce to get a better understanding of how the systems are linked. I showed him where to find leads, contacts, accounts, and opportunities exist in Salesforce. He asked, “where are the MQLs”?

He had heard the term as described by SiriusDecisions many times in his interview process and imagined it was another object in Salesforce. I then set about explaining that MQL was just a state of a lead. Leads can be “Recycled” “Unqualified” or of course “Marketing Qualified”. But the truth is an MQL doesn’t really exist in Salesforce, we marketers have to build it from scratch to track how many MQLs we create and pass on to sales team.

If I try running a Salesforce report on Lead status = MQL and created date = This Month, I come up with 4 leads, which is a good thing as it means my SDRs are quickly actioning and moving their leads and moving them along in our process.

My work around to be able to track how many MQLs are created in a given time period is to add system dates stamps and assign both Leads and Contacts that reach the point threshold of “MQL” to a Salesforce campaign. This enables me to track how many I’ve created in a certain time period and build nice views, dashboards, and reports.

The way it works is this.

In your salesforce create three new fields for leads

MQL First Date – make it a date field

MQL Most Recent Date – make it a date field

MQL Counter – make it a Number(18, 0) field

It should look like this:

Screen Shot 2015-08-26 at 5.06.21 PM

Do the same exact thing on the contact object

Screen Shot 2015-08-26 at 5.08.11 PM

Then let those fields synch over into Marketo. So that means get up, make coffee, bug your sales reps, whatever.

From here you’re going to adjust your lead lifecycle smart campaigns in Marketo to add data to these fields when a lead “MQLs.”  (Whats that? You don’t have a lifecycle smart campaign? more on this topic later)

Here is my smart list. The smart campaign is triggered either by the first “Lead is SFDC eligible” campaign, or by a change in lead score value.  I split the syncing lead to SFDC from updating the lead status to MQL as I want the lead to get into SFDC, get assigned a lead owner through my round robin rules, and then get back into Marketo with a lead owner.  That way when I send email alerts to the reps, it goes to the specific rep.

Screen Shot 2015-08-26 at 5.09.54 PM

And the flow looks like this. Notice I update the lead status only if the existing status is “Open, Target, Recycle, and MQL.”  This prevents Leads that have been disqualified by sales from turning MQL again.  The “MQL First Date” field gets written once and only one, hence the condition of if the field is empty update it.  Then for “MQL Most Recent”, that field can be updated over and over again.

Screen Shot 2015-08-26 at 5.10.21 PM

I can then run reports in SFDC on “MQL First Date” to show all leads that reached the point thereshold in a certain time frame as well as “MQL Most Recent Date” to show new and recycled leads!  The MQL counter shows how many times a lead has gone through your lifecycle flow. If the number is more than 3, that indicates a tire kicker, student, competitor, etc.

Let me know how you handle MQLs.

The Automation of Marketing Automation

Originally posted to commpro.biz

As a marketer, you’ve heard the legend of marketing automation. The tale goes that there’s software out there that can move prospects along the buyer’s journey, delivering highly personalized experiences, all with minimal effort from marketers. The story is a promising one, and one that many marketers have already invested substantial time and budget in, but it’s an incomplete one. Despite mass adoption, 85% of marketers don’t feel that they’re using automation to its full potential today[i].

The name marketing automation is, in itself, a bit of a misnomer. The word “automation” conjures up images of a solution that will robotically and effortlessly attract leads, nurture them, and surface the best prospects for Sales. In reality, marketing automation can do this, but only with a considerable amount of manual effort. This often requires a dedicated automation engineer or professional services agreement with the solution provider—both costly options and not at all automated in the true sense of the word.

Lead scoring—a core tenet of marketing automation—helps Sales and Marketing determine which leads are most qualified, the best ways to engage with them, and understand when they are ready to talk to sales. For established companies with a steady flow of inbound traffic and leads, a lead scoring system is absolutely necessary. Lead scoring, as you might have guessed, also requires a lot of implementation time to get right. I’ve personally built about five lead scoring models in my career and I will attest that each one took countless hours just to implement initially. Add to this the ongoing fine-tuning required to keep the scoring model in sync with business goals, the regular adjustments to campaigns and content, and the inevitable calls to customer support. The time required to keep automation running is substantial. But does it need to be this way?

At Fliptop, we believe that the future of automation is, in fact, automation. Leading marketers will take advantage of big data and predictive technologies to turn today’s manual, fairly static automation processes into dynamic, ever-evolving systems able to adjust marketing activity in real-time. Using predictive algorithms and tapping into the wealth of prospect data available (behavioral and demographic; native and public), automation systems will be able instinctively build and adjust lead scoring, routing rules, identify the next best marketing campaign to run, and tell marketers exactly where to invest their hard-fought budget. Marketing investments will become significantly more precise, allowing marketers to focus less on the science of marketing, and more on the art.

Most organizations are not quite there yet, but that doesn’t mean automation isn’t still powerful, game-changing software. It is, and the companies equipped to spend the time and resources required for proper automation maintenance are reaping the benefits. The truly innovative companies are supplementing these efforts with predictive technologies, able to look beyond traditional lead scoring rules, determine the characteristics of a good lead or account, and what it will take to convert them—resulting in huge gains in efficiency and sales.

According to Salesforce, marketing automation will be amongst the most piloted marketing software of 2015. This is great for companies and consumers because it will enable increasingly relevant, personalized experiences. Until automation can live up to the hype implied in their name, however, companies that seek truly automated marketing should always have predictive on the very same shortlist of must-have martech.

________________________________________________________________

[i] http://www.siriusdecisions.com/blog/eight-is-not-enough-increasing-adoption-of-marketing-automation-platforms/

 

How Netflix Knows: The Power of Predictive and What It Means for the Future of Marketing

originally posted to TargetMarketingMag.com

How well do you know your customers? Sure, you may have a grasp on some basic demographic information and order history, but how well do you really know them? Enough to deliver experiences tailored to their personal preferences? Enough to tell them which “Imaginative Time Travel Movie From the 1980s” they might enjoy next?

Netflix suggestions

Related story: Predicting Profits With Models

Even if you’re not in the business of streaming media, marketers across industries would do well to take a page out of Netflix’s book. The streaming giant has raised the bar on personalization, serving highly tailored recommendations based on any number of data signals — from viewing history to user behavior (browsing, scrolling and search patterns) to time, device and even location that a user is logging in from. The result is a new breed of picky consumer — ones who expect us to know exactly what they’re looking for … even before they do.

Using the data that surrounds our day-to-day marketing activity, and a little bit of math, leading companies are using predictive techniques to better serve prospects and sell more. And while the term “predictive” may conjure up images of a man behind a curtain, a Magic 8 Ball, of late night TV’s own Madame Cleo … magic it is not. Predictive marketing analyzes the data already contained within your native technologies — CRM, marketing automation, etc. — and data from across the public Web to identify and prioritize your top quality leads, accounts, campaigns and marketing activity.

Despite knowing just how influential a well-implemented personalization strategy can be, it’s still a tragically underutilized tactic and, according toone study, 71 percent of companies fail to personalize their Web experiences today. The reason is simple: Data is more abundant than ever, but the ability to process, analyze and derive actionable insights from this always-growing mountain of data is no small feat. While the Netflixes of the world may have the budget and headcount to dedicate to predictive personalization initiatives, SMBs are often left scratching their heads — overwhelmed and unsure of where to start.

In B-to-C, technology marketers can use most often provides Web users with helpful product or service recommendations. Personalization engines can look at a customer’s history to recommend products that algorithms dictate they’d be likely to buy. Based on a combination of past purchases, browsing history and any number of other factors, data is helping B-to-C companies surface the products and offers that are most likely to convert visitors into buyers.

In B-to-B, predictive generally takes the form of highly tailored educational content — targeted to prospects with attributes that indicate they have a high likelihood to buy. Based on how well the characteristics of a prospect align with where a company’s seen success in the past, B-to-B marketers can use predictive tech to engage prospects with content that they know is most likely to resonate. With the ability to know exactly how likely a prospect is to become a customer, sales benefits from the ability to prioritize and personalize follow-up to ensure that they’re addressing a specific prospect’s unique needs.

Netflix is but one of countless services that is transforming consumer preference and expectations. Amazon, Spotify, Facebook and LinkedIn are all examples of experience-focused companies, leveraging big data to predict what users want and tailor their experience in real time.