1. Library
  2. How Marketers Can Achieve Personalization at Scale, Part 2

How Marketers Can Achieve Personalization at Scale, Part 2

Light Mode
Achieving Meaningful Personalization Across Channels at Scale
What Metrics Should you Track
15 min

How do you meaningfully connect with what many consider to be the most skeptical audience? It’s a widely shared platitude that ‘developers don’t like being marketed to’. The truth is, most developers don’t like traditional marketing, but they can certainly be successfully marketed to.

They perceive traditional marketing as deceptive packaging, designed to pull the wool over their eyes. Their interest is in what lies beneath: the product, the surrounding community, and how quickly they can get back to what they really care about.

With developers, it becomes even more critical to make sure your messaging is on target. And while marketers are continuing to prioritize personalization, they are actually further behind than they think. VentureBeat published a research article on just how bad most marketers actually are at personalization.

To understand why most personalization efforts fail, lets first step back to examine what personalization actually means. Here’s Accenture’s definition:

“Personalization is the act of dynamically curating experiences to each individual and context in a seamless manner across channels. This includes all interaction, such as marketing, shopping and services experiences.” – Accenture Interactive

In simpler terms, it’s about delivering something ‘actually useful’. So why is this so difficult? Perhaps because as technology advances, it offers a double edged sword to marketers. It creates new channels that bring opportunities yes, but these opportunities force scrappy startup marketers to make tough decisions on where to focus and what to ignore.

It’s a no-brainer that message consistency and relevancy are critical to personalization. But if the average US user has 7.2 internet connected devices and the average marketer can only see 1 of them, it’s pretty obvious why even marketers with enormous budgets aren’t doing multi-channel personalization well.

And as if the marketers job wasn’t hard enough already, ADI research shows that ‘device switching’ is becoming more common. Users today often start a task on one device, such as online shopping, but continue it on another. This is particularly true with millennials (age 18 – 34) where 34% switch devices while shopping online.

I know I personally jump from my phone to my laptop when an app doesn’t support Apple Pay, or when I have to manually enter a lot of text. I’ve been burned many times by poorly built apps and ‘mobile-optimized’ websites losing my progress.

As Tamara Gaffney, principal analyst at ADI put it “it’s looking like a bit of a mess for marketers wanting to deliver the right messages at the right time to the right person.”

But there is a silver lining and it is that users have indicated that they actually want more personalization. According to the ‘State of the Connected Customer’ report, 63% of millennial users and 58% of GenX users are willing to share data with companies in exchange for personalized offers and discounts.

So how do you deliver a personalized experience when faced with a growing number of marketing channels combined with incredible device proliferation?

Part 1 of this guide focused on laying the right foundation by 1) defining actionable audience groups, 2) identifying data sources and 3) exploring data enrichment options. Part 2 examines ways of applying personalizations across channels and includes tips and resources to help you along the way.

Achieving Meaningful Personalization Across Channels at Scale

First identify the channels that matter: With so many channels available, how do you know which channels to use?

Start by listing out all of your customer touch points. For example:

  • Email marketing
  • Your website
  • Social media
  • Meetups & conferences
  • Webinars
  • Advertising on third-party sites

Each touch point offers numerous personalization opportunities. But marketers at startups have limited resources, so don’t try to hit them all. Choose 2-3 to start, run some experiments, and then move on.

Decide which channels are most important to you. That is 1) those channels that your highest quality customers prefer, and 2) those that perform well for you or your competitors. Bear in mind that channel performance varies based on the targeted audience segment and stage of the customer’s buying journey you’re going after.

Tools exist to help you understand which channels customers use and which perform well, but channel usage does not automatically imply preference and so it makes sense in some instances to simply ask your existing customers for their preferences.

Personalize on things that matter: Basic personalizations like merging in contact names may have been impressive at some point in time, but it’s table stakes now.

If the thing you’re personalizing doesn’t add value, it’s unlikely to deliver results.

It’s important to understand what really matters to your customers. A cookie cutter approach won’t work as user needs and preferences vary a lot. Building detailed and focused customer personas will come in handy here.

Broadly speaking, you have options to personalize based on demographic, behavioral, location and user supplied data. Let the insights you gain about your customers via persona research guide your decisions on what to personalize.

For example, your buyer persona research may reveal that a particular audience segment are running mission critical systems and looking for better system visibility. Performance & uptime matter a lot to them. If you’re offering a performance analytics service, you should make specific product recommendations based on analysis of their website i.e. “we noticed your website performance could be improved, and we think our product could help.”

The challenge is to find a way to do this well at scale i.e. programmatically, with minimal manual intervention. One approach is to segment users based on their stacks, and target the highest value segment with personalization and the others with more generic messaging. One interesting tool to help you with this stack analysis is Clearbit’s ‘technologies’ data. Send a domain through Clearbit and you’ll get a list of technologies the website uses.

Once you’ve identified your most important channels and the types of personalizations that matter to your customer, you should consider the strengths and limitations of each channel.

I’ve included a few channel examples below.


Email may be one of the older channels for personalization but it’s still unbeatable in terms of reach and ability to convert. Social media get’s all the attention these days, but research shows that there are over 4 billion email accounts worldwide, giving a reach 3x greater than Facebook and 15x greater than Twitter.

Email mostly falls into the middle of the buyers funnel so personalized tips and promotions can be used to drive retention and loyalty. As we gain more insight into who our customers are, we unlock increasingly exciting ways to surprise and delight them. Here’s an example.

VIP Appreciation
This email from Spotify targets fans of Charles Kelley and gives them early access to his solo tour tickets. The email works well because 1) it comes directly from Charles Kelley 2) it targets his top listeners and 3) rewards them for their loyalty and encourages them to continue listening.

The most common approach for delivering this segmented personalization is through drip or journey based automation. Just about every email marketing automation product offers this functionality, so take a look around your tool’s docs if you’re unfamiliar.

Fundamentally, this is how segmented drips work:

  1. Customer data is shared in real-time across your analytics tools, CRM, and email automation tool.
  2. A journey designer is used to segment customers using ‘and/or’ statements e.g. if Customer Location = San Francisco + Customer Purchase > $1000 then assign Customer to ‘VIP Segment # 4’
  3. Emails are crafted to target each segment e.g. send email about exclusive event in San Francisco to ‘VIP Segment # 4’
  4. Customers are automatically moved from one segment group to another based on the actions they take on your site, your app, or in previously sent emails.
  5. Various filters and delays organize and send emails to customers as they traverse your automations.

Here’s an example of Campaign Monitor’s Journey Planner:

Check out the Top Email Marketing Software Products (Capterra)
Read ‘9 emails every marketer should be sending’ (Campaign Monitor)


Mobile usage is definitely trending up. On the train, at the office or on the couch watching TV, consumers spend most of their waking hours on mobile. Often, developers will perform their research on a mobile device, and then switch to their computers to sign up, test tools and make the final purchase.

It’s therefore critical to modify your content to deliver the same compelling desktop experience on a smaller screen.

Deep Linking Apps

Mobile apps can be a great way to build an easy to access, on-the-go experience for users but retention is not easy. The average app loses 77% of its users in the first three days.

Deep linking can be used to boost retention, by driving users from a mobile browser directly into the relevant page in your app.

For example, a SaaS provider that emails usage reports to customers should use a deep link to take them directly to their in-app reporting section where they can view details and run additional reports. If you don’t have a mobile app, you should make sure that your users land where they expect on your mobile site. There’s little more frustrating than starting from zero after clicking a direct link.


Outside of your organic followers and fans, social media engagement isn’t free. To get the most value for your money, you have to personalize using the advanced targeting tools that Facebook and Twitter offer.


Most social channels offer re-targeting options where you can serve tailored messages based on a user’s historical behavior.

For example, you can use Facebook ads to retarget users who visited your website. Setting up retargeting is quite simple. You install a snippet of code on your website that cookies users who visit your website. As you’re experimenting with different channels, you might consider implementing Segment to allow you to quickly test different services without having to manually push new code.

Each tracking code will send information back to the top level network about which pages were viewed on your site by a particular visitor. You can then, for example, use Facebook’s Custom Audience preferences to set up specific retargeting ads.

Are you seeing a large number of visitors view a particular pricing page on your website but not make a purchase? Have Facebook present an ad to those exact people with a limited time promotion to get them to come back and convert.

Pre-existing communities

In addition to social targeting tools, you should also find and engage with micro-communities and supernodes within your segment. To be fair, this doesn’t scale quite as well due to the limited size and number of these groups, but they provide lasting value outside of any individual product signups.

How do you find these micro-communities? Subreddits, Facebook Groups, Twitter Influencers & existing Slack teams are a great starting point.

When you find a community or supernode that’s clearly full of qualified folks, join them, participate in the group and learn from them. Use what you learn to refine your personas and the campaigns in other channels, and then build custom personalized campaigns that target these micro-communities specifically.

Read ‘The Rise of Micro-communities in 2016’ (Social Media Weekly)
Download the ‘Definitive guide to social media’ (Marketo)
Read ‘How to Network With Facebook Groups’ (Social Media Examiner)
Read about ‘Secret Facebook Groups’ (AdEspresso)
Check out this handpicked selection of ‘Top Slack Communities

What Metrics Should you Track

As always, look beyond vanity metrics such as visitors, subscribers & followers. Here’s what you should be tracking for each channel:

Email: In addition to open & click rate, it’s worthwhile tracking:

  • Engagement over time to understand the best times to send emails.
  • Unsubscribe rates tied to specific topics.
  • Device type, to inform decisions on your designs as well as content length.
  • Email client, to ensure you’re optimizing for the right clients.

Social: Generally, for social you should track:

  • Audience growth rate to understand your social momentum.
  • Engagement metrics after-click, i.e. visit time and visitor return frequency.
  • Conversion rate of social inbound visitors to signups.

As in most things at early-stage startups, you should be very flexible to change. You should be constantly experimenting to find new areas for growth, and should fight the urge to sit still when you’ve found your first bit of success. As your company grows, the tactics, channels, and metrics you used yesterday may not be applicable today, and only flexible teams are capable of not getting stuck.

Achieving meaningful personalization isn’t an overnight process, especially if you’re targeting the skeptical developer community. The good news is that it’s not an overnight process for your competitors either, so use this guide to get to work as soon as you can.

Thanks for reading. Let me know on Twitter if you have anything else to add, and sign up for Heavybit Updates if you’d like to get our Library content directly to your inbox.