One size does not fit all

Segmentation is the difference between talking directly to an audience that’s interested in what you’re selling and standing on the street corner with a megaphone. It always has been. But the rise of new digital channels, new data, and new tactics have taken this segmentation thing to a whole new level.

Today, people-based marketers don’t treat segmentation like an annual planning exercise. They use it on a daily basis to target their audiences with a laser-focus and level of relevance that would make a sales person blush.

Which is great. Because it makes a lot of important things possible:

• More efficient, cost-effective campaigns that divert media spend away from people who’d never buy in the first place.
• Consistently relevant customer experiences that span different channels like digital ads, emails, and websites.
• Highly relevant content, offers, and recommendations that convert more and annoy less.

The key is identity resolution

How you segment your audience will depend on what you’re trying to do. But the foundation of every successful people-based segmentation strategy is data. Data that’s cleansed, updated and crucially, ties back to real people.

That calls for identity resolution – the ability to tie 1st, 2nd, 3rd party, behavioural and contextual data back to real people in a privacy-safe way.

Without it, both segmentation and targeting suffer.

First, identity resolution gives you a clearer picture of your audiences. It lets you analyse customer and prospect activity across touchpoints. So you aren’t building segments based on misleading fragments of data.

Then it allows you to activate all that people-based data across all your digital touchpoints and platforms. So everything you do with Vendors A, B and C ties back to the same prospects and customers.

Now all the cool, dynamic display, social, search, video and mobile ad units you’re paying for can actually be used to their full potential.

We’ll discuss identity resolution some more a little later in the eBook.

But first let’s take a look at how some of the most sophisticated marketing teams are using identity resolution to power a whole range of effective segmentation and targeting tactics.

1. Audience Suppression

Audience suppression is all about making sure certain segments aren’t exposed to certain ads.

It may sound counter-intuitive (because what advertiser wouldn’t want to, you know, advertise). But it makes a lot of sense in a number of situations.

• When the offer is irrelevant: You wouldn’t offer 15% off to a customer already getting a 20% discount.
• When they’ve already converted: You wouldn’t offer a great acquisition deal to a loyal, returning customer when you don’t need to.
• Or when they have an open customer service ticket: And you want to avoid overkill.

Why it's a smart move

When you can accurately segment your audience into lists like ‘people who’ve already converted’ and ‘people talking to sales agents’, you can avoid a lot of wasted media spend on misplaced offers.

Also, by suppressing the right segments, you can actually increase your conversion rate. Because you can make sure it isn’t skewed by people who’d never have converted in the first place.

Even better, you can make sure those people don’t suffer the ill-effects of brand fatigue (and plain old annoyance).

Who's using it?

Widely used in retail, one great example of audience suppression actually comes from politics.

Here, parties used publicly available lists of ‘early voters’ (people who vote by mail) to make sure they were only targeting people who hadn’t voted yet.

Considering the fact that these lists contain hundreds of thousands of names, they avoided a huge amount in wasted media spend.

Data Onboarding

Data onboarding is an identity resolution process to match your offline customer data to your online marketing platforms. And it’s a big deal when it comes to effective segmentation.

Because it connects your offline customer data (like in-store purchase data, call centre data and demographics) to anonymised online identifiers (like online devices and digital IDs) in a privacy-safe way.

That means you can activate all your segments in the ad platforms, DMPs and social channels where you run your campaigns.

2. Look-alike modelling

Look-alike modelling is all about understanding the key attributes of a small subset of your most active customers.

By understanding these attributes, you can learn a lot about what makes these customers tick. But more importantly, you can then target other people who fit the same bill.

It takes some data wrangling. But with ‘double and triple the results of standard targeting[2]’, it’s a tactic worth adopting.

Why it's a smart move

Instead of looking for prospects and lists by demographic attributes like gender and age only, look-alike modeling gives you the chance to get a lot more specific by including psychographic attributes and behavioural data too.

For instance, if your best customers turn out to be college students who read the Wall Street Journal and buy Red Bull by the 12-pack, then it makes sense that your targeting efforts shouldn’t be restricted to college students alone.

By leveraging your offline customer data and third-party data enrichment vendors, you can build incredibly specific look-alike models and target those who are most likely to buy.

Who's using it?

A leading bank wanted to expand its direct marketing campaigns to reach prospects online.

So they turned to data onboarding and look-alike modelling to find prospects that fit the criteria of their best customers. The results were impressive: 150% improvement in response rates and a massive 40% improvement in ROI.

3. Segment-based retargeting

The most common form of retargeting is a function of cart abandonment, where the pair of shoes you nearly bought (but then didn’t) follow you around the internet.

In this case, the only contextual data available to those ad units is the fact that you were looking at a certain product or product range. While that’s a great place to start, you can do a whole lot more when you add data about your segments.

Why it's a smart move

By tying your retargeting tactics to your segmentation strategy you can actually go a whole lot further than productcentric retargeting. You can retarget female customers from Estonia who bought summer shoes for their winter break in 2014.

That is, you can use retargeting ad units to target people at different stages in the customer decision journey, depending on the kind of behaviour they’ve demonstrated on your site and other platforms too.

If a certain segment of yours has a higher propensity to buy during certain seasons, then your retargeting ads can dynamically deliver certain customised messages to that segment, while targeting others with more general acquisition offers.

Who's using it?

Even though retargeting is most prominent in the retail industry, it has been successful in other verticals. For instance, one niche insurance vendor that offered online quotes in real-time (as opposed to over the phone) used retargeting to make sure smallbusiness decision-makers completed the request-a-quote form online.

Not only did users who left the form midway come back after being retargeted, a meaningful proportion of them actually ended up converting into sales.

CRM retargeting

CRM retargeting is all about using the data stored in your offline CRM systems (purchase history, contract expiration dates...the good stuff) to retarget your customers online.

And it makes sense for any brand with a CRM system and customers – so, pretty much anyone reading this sentence.

First, because sending acquisition deals to customers is a waste of media spend and just plain ineffective. And second, because as any salesperson worth their salt will tell you, you’re a whole lot more likely to sell to someone you’ve already sold to before.

But even beyond retargeting, your offline CRM data can give you a lot of important insight about buying patterns and trends. So it’s worth using it to fuel a lot of the tactics we’re writing about in this eBook such as audience suppression, look-alike modelling, segmenting your loyalty tiers, and reactivation.

4. Segmenting loyalty tiers

If you think about it, marketers have used loyalty tiers to segment their customers for years. From tiered credit cards to frequent flyer programmes, it’s been a good idea to have ‘gold’ and ‘platinum’ members for some time now.

So when it comes to digital marketing, targeting those tiers with specific offers is really just low-hanging fruit. And the smartest digital marketers are already picking it.

Why it's a smart move

By onboarding your offline customer data and segmenting it into different loyalty tiers, you can target those customers with unique offers online. You can also optimise your bidding strategy for audiences that are likely to convert at a higher rate and spend more.

Your offers become more specific, you maximise the value of your dynamic digital ad units, publishers, and platforms and most important of all – your most loyal customers get exclusive experiences.

Who's using it?

As you’d imagine, this is a tactic most popular among industries where loyalty tiers already play a big role, such as financial services, retail, and travel.

But even if you don’t already have a tierbased loyalty program already set up, it makes a lot of sense to build one around your offline customer data.

One of the most famous uses of loyalty tiers is Starbucks Rewards where customers get better offers the more they use Starbucks’ mobile application. And the more information they share, the better the offers get. For instance, gold card members get treats and drinks for free on their birthdays.

5. Dynamic creative and site optimisation

Home pages, emails, and ad units have all evolved to become dynamic containers capable of assembling unique combinations of content from multiple different creative elements.

Back them up with a smart segmentation program and you have a living, breathing campaign that can deliver exactly the right experience to each consumer.

Why it's a smart move

In its most basic form, dynamic creative is used to make sure some people see one message and other people see another. But the more segment data you can activate, the more sophisticated your campaigns can become.

You may start with separate home pages for prospects and customers. But once you include data on things like their purchase history and seasonal preferences, your ability to make highly customised offers starts to multiply.

For instance, you could target golf-lovers and tennis-aficionados separately with the same dynamic ad unit. Going one step further, you could even target male golf lovers in Australia and female golf lovers in the U.S. with different creatives depending on how sunny it’s going to be next weekend.

Your customers get more relevant experiences and you gain a platform for testing more variants and options.

Who's using it?

While retailers have been quickest to adopt dynamic creative technology, their popularity is rapidly growing across other industries. For instance, we’ve seen a technology company run as many as 50 home page variants that serve 50 different segments based on their email database.

In another case, a retailer in the beauty industry uses dynamic landing pages to showcase either their skincare or cosmetics range based on the customer’s preferences.

The importance of identity resolution

Some marketing interactions involve a consumer’s known identity, such as sending an email. Other interactions, like displaying an online ad, happen when the consumer is anonymous.

Most marketers struggle to connect the dots across the customer journey and lack a comprehensive, unified view of it.

In the context of segmentation, that’s a huge loss. Because it means you can’t create audience segments based on a clear understanding of past interactions. Even worse, you can’t then activate that segment data and target those anonymous users online.

The key to squaring this circle lies in anonymising and accurately linking data about online and offline interactions across channels and devices in a secure and privacy-safe way. That’s identity resolution.

We’ll stop short of tooting our own horn here, but if that’s a problem you need solving, this is a good place to start.

6. Targeting based on anonymous online activity

Just because so many digital interactions are anonymous doesn’t mean you can’t use them to create actionable segments.

An anonymous visitor who trawls the sunglasses’ section of your website is still, clearly, in your sunglass-interested segment. You don’t know who they are but you know their behaviour and their browser.

Why it's a smart move

Used intelligently, behaviour-based segments can lead to super-relevant marketing messages. You don’t need to wait for someone to fill out a form to reach them with more relevant stories wherever you interact with them.

For instance, if someone on an electronics website spent half hour on a page about a certain television, it makes a lot of sense to start targeting them as someone interested in that television.

So once you link their anonymous browsing data to your anonymous targeting data, you can start sending them more targeted acquisition deals and complimentary offers for things like HDMI cables and gaming consoles.

Who's using it?

In the car industry, one brand uses anonymous data from its display campaigns and site traffic to determine which cars someone’s interested in and how high their propensity to buy actually is.

By doing this, they can deliver unique homepage options to different users (making the car they’re interested in the hero) and analyse which stage of the journey a user is in. Based on all of this, they then programmatically determine which messages and offers should be made next.

7. Geo-targeting

Geo-targeting has gone from digital marketer’s fantasy to tangible reality in a very short time.

The age-old fantasy of ‘man walks into store and store offers man discount’ may not be as common as marketers had once expected, but there are a number of emerging ways to use location data to customise offers.

Why it's a smart move

At the very least, geo-targeting is good customer experience hygiene. It makes sure home pages are in the right language, prices are in the right currency, and whether or not products are available in your country.

Taken a few steps further, geo-targeting can be used to personalise messages and even push certain offers depending on where someone is within a brick-and-mortar store.

Who's using it?

Use of geo-targeting as an effective segmentation tactic is still in its infancy, but there are some interesting uses for the technology. For instance, one hospitality brand uses geo-targeted offers to welcome customers when they check-in and alert them to offers to local activities and stores.

8. Segmenting based on offline purchase behaviour

For all the noise and interest generated by eCommerce and digital marketing, the vast majority of retail purchases still happen in store. Which means the vast majority of useful behavioural data on your customers is stored offline.

So another smart move for the segmentation-curious marketer is to tap into the wealth of in-store purchasing data you have stored offline.

Why it's a smart move

Even though most of your efforts to identify, attract, and persuade your prospects happen online, the proof of that pudding lives offline. By tapping into your offline transactional data, you can create segmentations based on the products that were actually sold the most.

That means being able to connect the dots between your online activities and their offline impact. For starters, this makes for highly specific targeting opportunities in the context of other tactics like audience suppression and look-alike modelling.

But on a more macro level, this gives you the ability to spot buying patterns in the data and inform the rest of your marketing.

Another valuable use of offline data is to connect your call centre data with your digital ad units to either suppress certain segments or give them special offers.

Who's using it?

Brands in nearly every industry are using data onboarding to activate their offline transactional data.

One particularly interesting use case is in the automotive industry where offline data about car purchases is used to inform the targeting activities of insurance products.

9. Segment-based cross-selling and upselling

Smart cross-selling and upselling are essential to effective digital marketing.

But when the only context you have to consider is the product page someone’s on, your ability to make a relevant recommendation is limited. So when you throw into the mix their purchase history and behavior data, you can start to make more relevant recommendations based on the segment they belong to.

Why it's a smart move

For one thing, a segment-based view of recommendations would inform your strategies across channels. That is, if someone bought a product on your website, it isn’t enough to make the right recommendation on your website.

The smart thing to do is to advertise the complimentary product to them through your search, mobile, social, and display channels too.

Even better, you can understand how customers use different channels and tailor your messaging to match. For instance, if they use mobile to browse, then you can run brand recognition campaigns on mobile and more specific product recommendations when they’re back on their desktops.

The missing link, of course, is identity resolution (see page 3). It makes a huge difference to both your segmentation and cross-sell strategies.

Another benefit of segment-based recommendations is that the feedback loop created by testing offers and measuring responses can then be used to inform the rest of your segmentation tactics.

Who’s using it?

One great example of smart segmentbased upselling can be found being used by large banks. Here, customers who’ve recently bought a credit card are targeted with offers on insurance products depending on the kinds of purchases they’ve made most recently.

10. Reactivation

Reactivation is all about identifying the customers who haven’t engaged with you in a while and targeting them with unique, relevant offers.

Instead of ignoring this segment like a foregone conclusion, it makes sense to gauge when the right time to re-engage is and find out what offers really work.

Why it's a smart move

It makes a lot of sense to treat dormant customers like a separate segment. On the one hand, they are former customers so they’re likelier to reconvert. But on the other, getting them to convert again can be even harder than pursuing net new customers.

strong>Who's using it?

A number of really smart marketers are using reactivation tactics in interesting ways. In subscription-based businesses like magazines and publishing, it’s been proven that it’s a lot easier to get people to re-engage with content, than it is to engage new prospects.

Another great case for reactivation is in niche holiday resorts where customers may not come back for several years at a time. A smart reactivation strategy here can be the difference between repeat, high-value business and a forgotten story in a Facebook album.

Smart marketing doesn't have to be hard

It’s funny how after years of innovation and digitisation, one of marketing’s most powerful tools may be the oldest trick in its book.

Segmentation and targeting has been around for decades. But it’s only in the last few years that data and technology have come together to make it what it is today.

Intelligent, proactive, and highly conducive to testing, effective segmentation in the digital age is crucial to marketing success.

That being said, it’s important to remember that at the end of the day, the success of any segmentation tactic comes down to your ability to tie all your data back to real people. In particular, as we’ve discussed, offline data holds the key to effective segmentation.

Connect to, understand and then activate your offline data across all your digital channels, and there’s very little you won’t be able to accomplish.

The technology exists, the data exists, and the willingness to deliver more targeted marketing exists – smart segmentation isn’t just a tactic.

It’s a people-based marketer’s imperative.