How can we help you?
We want to hear from you. And it’s easy to reach us. Give us a shout at 844-886-2252, send us an email at hello@mediumgiant.co, visit our contact page, or fill out the form right here. You can expect a response within two business days.
What are you looking for?
A step-by-step guide to using analytics to drive more customers to your business
“We do analytics.”
Companies say this all the time. But what they mean by “analytics” is not what we mean.
Every company gathers data (hopefully), but not every company acts on that data to optimize its advertising as well as it could. In fact, marketing analytics are only influencing 53% of decisions, according to Gartner.
The problem is the word “analytics” has become synonymous with machine learning or something like cryptocurrency. The meaning of the word has gotten distorted across the marketing landscape.
When we say “analytics,” we mean using a scientific method to drive more customers — and therefore revenue — to your business. This involves creating a hypothesis based on data, experimenting, and iterating.
Analytics is a constant cycle of answering, measuring, and optimizing for these two questions:
- What has happened?
- What do we think is going to happen?
After we deploy media toward our hypothesis, we have new data to ask that first question again. This flywheel of gathering data, using data to hypothesize, then deploying, is the cornerstone of analytics.
Customer segmentation is a core component of the misused term “analytics.” The following gives an overview of our curiosity-driven approach of using data and insights to drive better results for your next media campaign.
The 4 elements of customer segmentation
If you aren’t using data to guide your hypothesis, then you are shooting in the dark.
We seek to answer these four questions before deploying medias:
- Whom do you want to reach?
- Where are they by geographic area?
- How are you going to reach them?
- What is the message?
No. 1: Whom do you want to reach?
When most people think of “target audience,” they go straight to demographics. Although demographics are important, the real key is intent to buy.
Demographic segmentation doesn’t give you a clear picture of your buyers.
If you were to ask 100 women ages 22–24, all in the same economic bracket, if they owned a Louis Vuitton handbag or a Michael Kors handbag, you would get different answers.
We’re always trying to get companies to understand that we’re looking for behavior, not demographics. This behavior is what we incentivize with media. We are trying to get someone to buy your product or service, after all.
If you sell a Chevy Trailblazer, it’s not enough to just find people who want to buy cars or even a Chevy Trailblazer. You have to find someone who wants to buy it from your dealership.
We begin by looking for people who have already taken that desired action. First- and third-party data is integral in finding this.
Your customer data (first-party) is a great starting point to begin creating an image of your targets. We then augment and update this data with third-party data to crystallize that persona even better. It helps us answer key questions: Who is going to your competitors? Who went to your store but left? Where are they coming from?
Once you have used data to find customers with intent to buy, you create look-alikes for this segmented audience.
No. 2: Where is your audience?
Geography data is an underappreciated source of targeting info, especially for national or e-commerce brands. We’ve found that combining demographic data with ultratargeted geographic data gives you a far better picture of the people you should be marketing to.
Once we’ve done some basic analysis, which largely comes from the customer database, we augment with third-party data that tells us residential data (latitude and longitude) of people who leave their house and go to your parking lot or the parking lots of your competitors. From this data we can find out the preferences of the marketplace by area.
This helps us figure out market share and engagement in the category. You’ll never be able to see what’s happening with your competitors, but we can.
Our method involves analysis of ZIP codes in a target county, state, region, or nation. We make a map using current customer data to see the ZIP codes for which your brand has good market penetration.
We then use our look-alikes to create a visual map of all the ZIP codes we believe your audience resides in. This comparison uncovers strategic opportunities to either double down in places that are working, or invest in new areas where your audience is, but market penetration is low.
It’s easy to understand how geographic data helps physical locations, but how does it make sense for large or national brands?
For distributed companies such as insurance or e-commerce, their audiences are more targeted by geography than people think. At a meta level, if you look at the biggest e-commerce brand in the world, Amazon, even it has not penetrated evenly across the nation. There are ZIP codes that spend way more than others — and way differently.
It makes sense that Amazon’s customer data by ZIP code could deliver very important distinctions. Understanding order value by ZIP code and the profit margins of shipping by ZIP code will influence how Amazon advertises to different segments of the country.
Where your audience is helps you uncover latent opportunities to achieve far greater results from your marketing.
No. 3 How are you going to reach them?
The media mix you deliver to your target audience is how you reach them. You might even have a couple of different media stacks based on your different customer personas.
There are a couple of ways advertisers create an initial media mix:
- Discover the consumers’ media preferences.
- Gather data on what large-spending advertisers are doing and assume they are making more intelligent decisions.
From testing, we’ve learned that the latter is the more valuable way to create an initial mix. After we deploy and have our own data, obviously we tweak and optimize to create our own unique mix that fits the company.
Learning consumers’ media preferences is difficult and doesn’t always show results. Just because a person reads the newspaper doesn’t mean they want to see an ad for a Chevy in that newspaper.
Generally speaking, larger spenders are more intelligent with their media purchases. There are multiple data sources to understand advertiser data by vertical among your competitors and large spenders to help build this media mix.
This, combined with historical first-party data of what has worked, is incredibly valuable in determining how you can reach your audience.
No. 4: What is the message?
Your creative delivers the message that (hopefully) turns your targets into customers. Your customer personas should inform your creative by using language and imagery that resonate with that target audience.
The key is to have multiple creative sets that you can A/B test. Creative testing should never end. It’s always a work in progress to find the media that does the best job at converting customers. Even when you find something works, you need to consider a refresh so people don’t get burnt out seeing the same ad again and again.
After you deploy your creative, you have to test your hypothesis. “We think this is going to work. Let’s see if it really does.”
After two weeks, it’s time to look at data and compare it with the transaction file. What incremental new customers have come into that location or business? We know where we deployed; now we’re seeing activity. Let’s match that up.
Once we test, we optimize. When we do, we usually get an 8%–14% boost in performance.
The 3 types of customer types
Most customer databases are just a sliver of the total market.
Companies hover around 8%–10% fractional market share. When we build the robust ecosystem of customer segmentation, you see that you’re losing a major share in your areas.
We group people into three customer types:
- People who prefer you to competitors.
- People at the margin who can be swayed either way.
- People engaged in your category but who don’t want to buy your product or service.
The most important questions when advertising: Who is likely to take the desired outcome with you as the provider, and where are they likely to come from?
The people in the first category are a no-brainer. You should double and triple down on people who prefer your product, service, or business.
Category two is where you’ll find a large group of people who need to be swayed one way or the other. Your marketing is going to get the job done for them.
The third category is the rest of the market: They may be interested in what you sell but not necessarily interested in you. We like to stick with the first two to increase our market penetration before moving to the third group.
When we have access to first-party data and overlay with third-party data, we get a slightly more crystallized version of your location plus a competitive marketplace set.
Your set is just a drop in the bucket of the total demand in your unique trade area, let alone the entire tableau. Breaking down your audience into these three categories can help you understand whom to market to and what message will resonate.
Start where you are gaining traction
Many advertisers see all the new opportunities to advertise and wonder where to start. We always recommend starting with ZIP codes where you are currently getting traction. Find out where the latent or unmet demand is in that geography. Establish beachheads where you can incise and deepen your existing penetration.
If you find that you have multiple areas that show the same market penetration and audience size, there is no data-driven way yet. Once we have our first set, we can revisit.
Once you’ve established this, we tactically expand to new geographies while always testing along the way.
The scientific method of customer segmentation
An advertiser’s job when using analytics is to always be answering these two questions:
- What has happened?
- What do we think is going to happen?
Once we deploy, the job becomes tweaking and optimizing based on a number of factors:
- Creative
- Media channel
- Cadence
- Year considerations
After three to six months, you’ll have a playbook. Here is what works, in what media channels, for this audience, in these geographies. And everything has been optimized.
This method of customer segmentation is the core component of using analytics to drive more customers to your business.
A step-by-step guide to using analytics to drive more customers to your business
“We do analytics.”
Companies say this all the time. But what they mean by “analytics” is not what we mean.
Every company gathers data (hopefully), but not every company acts on that data to optimize its advertising as well as it could. In fact, marketing analytics are only influencing 53% of decisions, according to Gartner.
The problem is the word “analytics” has become synonymous with machine learning or something like cryptocurrency. The meaning of the word has gotten distorted across the marketing landscape.
When we say “analytics,” we mean using a scientific method to drive more customers — and therefore revenue — to your business. This involves creating a hypothesis based on data, experimenting, and iterating.
Analytics is a constant cycle of answering, measuring, and optimizing for these two questions:
- What has happened?
- What do we think is going to happen?
After we deploy media toward our hypothesis, we have new data to ask that first question again. This flywheel of gathering data, using data to hypothesize, then deploying, is the cornerstone of analytics.
Customer segmentation is a core component of the misused term “analytics.” The following gives an overview of our curiosity-driven approach of using data and insights to drive better results for your next media campaign.
The 4 elements of customer segmentation
If you aren’t using data to guide your hypothesis, then you are shooting in the dark.
We seek to answer these four questions before deploying medias:
- Whom do you want to reach?
- Where are they by geographic area?
- How are you going to reach them?
- What is the message?
No. 1: Whom do you want to reach?
When most people think of “target audience,” they go straight to demographics. Although demographics are important, the real key is intent to buy.
Demographic segmentation doesn’t give you a clear picture of your buyers.
If you were to ask 100 women ages 22–24, all in the same economic bracket, if they owned a Louis Vuitton handbag or a Michael Kors handbag, you would get different answers.
We’re always trying to get companies to understand that we’re looking for behavior, not demographics. This behavior is what we incentivize with media. We are trying to get someone to buy your product or service, after all.
If you sell a Chevy Trailblazer, it’s not enough to just find people who want to buy cars or even a Chevy Trailblazer. You have to find someone who wants to buy it from your dealership.
We begin by looking for people who have already taken that desired action. First- and third-party data is integral in finding this.
Your customer data (first-party) is a great starting point to begin creating an image of your targets. We then augment and update this data with third-party data to crystallize that persona even better. It helps us answer key questions: Who is going to your competitors? Who went to your store but left? Where are they coming from?
Once you have used data to find customers with intent to buy, you create look-alikes for this segmented audience.
No. 2: Where is your audience?
Geography data is an underappreciated source of targeting info, especially for national or e-commerce brands. We’ve found that combining demographic data with ultratargeted geographic data gives you a far better picture of the people you should be marketing to.
Once we’ve done some basic analysis, which largely comes from the customer database, we augment with third-party data that tells us residential data (latitude and longitude) of people who leave their house and go to your parking lot or the parking lots of your competitors. From this data we can find out the preferences of the marketplace by area.
This helps us figure out market share and engagement in the category. You’ll never be able to see what’s happening with your competitors, but we can.
Our method involves analysis of ZIP codes in a target county, state, region, or nation. We make a map using current customer data to see the ZIP codes for which your brand has good market penetration.
We then use our look-alikes to create a visual map of all the ZIP codes we believe your audience resides in. This comparison uncovers strategic opportunities to either double down in places that are working, or invest in new areas where your audience is, but market penetration is low.
It’s easy to understand how geographic data helps physical locations, but how does it make sense for large or national brands?
For distributed companies such as insurance or e-commerce, their audiences are more targeted by geography than people think. At a meta level, if you look at the biggest e-commerce brand in the world, Amazon, even it has not penetrated evenly across the nation. There are ZIP codes that spend way more than others — and way differently.
It makes sense that Amazon’s customer data by ZIP code could deliver very important distinctions. Understanding order value by ZIP code and the profit margins of shipping by ZIP code will influence how Amazon advertises to different segments of the country.
Where your audience is helps you uncover latent opportunities to achieve far greater results from your marketing.
No. 3 How are you going to reach them?
The media mix you deliver to your target audience is how you reach them. You might even have a couple of different media stacks based on your different customer personas.
There are a couple of ways advertisers create an initial media mix:
- Discover the consumers’ media preferences.
- Gather data on what large-spending advertisers are doing and assume they are making more intelligent decisions.
From testing, we’ve learned that the latter is the more valuable way to create an initial mix. After we deploy and have our own data, obviously we tweak and optimize to create our own unique mix that fits the company.
Learning consumers’ media preferences is difficult and doesn’t always show results. Just because a person reads the newspaper doesn’t mean they want to see an ad for a Chevy in that newspaper.
Generally speaking, larger spenders are more intelligent with their media purchases. There are multiple data sources to understand advertiser data by vertical among your competitors and large spenders to help build this media mix.
This, combined with historical first-party data of what has worked, is incredibly valuable in determining how you can reach your audience.
No. 4: What is the message?
Your creative delivers the message that (hopefully) turns your targets into customers. Your customer personas should inform your creative by using language and imagery that resonate with that target audience.
The key is to have multiple creative sets that you can A/B test. Creative testing should never end. It’s always a work in progress to find the media that does the best job at converting customers. Even when you find something works, you need to consider a refresh so people don’t get burnt out seeing the same ad again and again.
After you deploy your creative, you have to test your hypothesis. “We think this is going to work. Let’s see if it really does.”
After two weeks, it’s time to look at data and compare it with the transaction file. What incremental new customers have come into that location or business? We know where we deployed; now we’re seeing activity. Let’s match that up.
Once we test, we optimize. When we do, we usually get an 8%–14% boost in performance.
The 3 types of customer types
Most customer databases are just a sliver of the total market.
Companies hover around 8%–10% fractional market share. When we build the robust ecosystem of customer segmentation, you see that you’re losing a major share in your areas.
We group people into three customer types:
- People who prefer you to competitors.
- People at the margin who can be swayed either way.
- People engaged in your category but who don’t want to buy your product or service.
The most important questions when advertising: Who is likely to take the desired outcome with you as the provider, and where are they likely to come from?
The people in the first category are a no-brainer. You should double and triple down on people who prefer your product, service, or business.
Category two is where you’ll find a large group of people who need to be swayed one way or the other. Your marketing is going to get the job done for them.
The third category is the rest of the market: They may be interested in what you sell but not necessarily interested in you. We like to stick with the first two to increase our market penetration before moving to the third group.
When we have access to first-party data and overlay with third-party data, we get a slightly more crystallized version of your location plus a competitive marketplace set.
Your set is just a drop in the bucket of the total demand in your unique trade area, let alone the entire tableau. Breaking down your audience into these three categories can help you understand whom to market to and what message will resonate.
Start where you are gaining traction
Many advertisers see all the new opportunities to advertise and wonder where to start. We always recommend starting with ZIP codes where you are currently getting traction. Find out where the latent or unmet demand is in that geography. Establish beachheads where you can incise and deepen your existing penetration.
If you find that you have multiple areas that show the same market penetration and audience size, there is no data-driven way yet. Once we have our first set, we can revisit.
Once you’ve established this, we tactically expand to new geographies while always testing along the way.
The scientific method of customer segmentation
An advertiser’s job when using analytics is to always be answering these two questions:
- What has happened?
- What do we think is going to happen?
Once we deploy, the job becomes tweaking and optimizing based on a number of factors:
- Creative
- Media channel
- Cadence
- Year considerations
After three to six months, you’ll have a playbook. Here is what works, in what media channels, for this audience, in these geographies. And everything has been optimized.
This method of customer segmentation is the core component of using analytics to drive more customers to your business.
More Like This
