Too many marketing consultants build buyer and user journeys with anecdotal data: one-off stories they glean from prospect and client interviews that they feel match up with the truth. These days, it’s inexcusable because we have fantastic user data that we can use to build fantastic journeys. If you want to learn how to create meaty, data-driven buyer journeys, read on!
Buyer Journey Definition
For the purpose of this article, we’re examining marketing’s impact on the buyer’s journey up to the point of sale. In today’s world, buyers frequently have their mind made up by the time they get to sales because they’ve done all their research online before speaking to a salesperson. A standard buyer’s journey with this definition will include an awareness stage, consideration stage, and decision stage or you could use the older AIDA model. More detailed buyer’s journeys may include more stages with extra detail. Some marketers will draw this out as a horizontal scale, others will draw it as a funnel. Personally, I like the funnel style because it fits text descriptions more conveniently. This article will focus on how to make the buyer’s journey more data-driven. It will not cover the basics.
What data belongs in a Buyer’s Journey?
If you’ve already got a buyer’s journey complete and now you need to layer on the data to back it up, your buyer’s journey is wrong and you need to start again. Anyone can start with a statement and find the stats to back them up, but that’s how you get faulty insights. You have to start with the data first and do quantitative analysis, then layer on qualitative context and then insights. That said, statistical analysis isn’t usually included in a visual buyer’s journey, it’s usually in the PowerPoint speaker’s notes or in a written report that goes along with it. You present the graphics and feature some stats, but you keep most of the stats in the final report or as backup in case a VP calls you out. How customers become aware of your company and products should be supported by data. You should already have this from Google Analytics and/or your marketing automation platform.
How does marketing drive awareness?
Create a pie chart of how prospects become aware of your company. What was their first marketing touchpoint? This is important because the lion’s share of marketing’s budget and time in many companies goes into awareness and getting new clients. Separate Organic traffic into branded and non-branded Here, it can be useful to attempt to separate organic traffic into branded (people explicitly searching for your brand) and unbranded. Branded traffic is traffic that gained awareness of your company and products through another channel, while unbranded traffic may be prospects who didn’t know about your company before making their search. Check Google Search Console to see what portion of traffic is branded and whether any specific page takes all the branded traffic. Some companies are lucky with this and get all their branded traffic to the homepage, while sub-pages are almost all unbranded. This can make it a LOT easier to separate branded vs non-branded traffic. Non branded traffic can be improved through SEO, while branded traffic is usually coming in from other marketing activities. If you’re bidding on company brand keywords or they’re showing up in paid search, you may want to segment these off as well. This is way easier because you get keywords data with Google Ads. Add in tracking where it’s missing This is the wrong time to fix tracking, but in any case, you’ll need email and offline sources tracked as best as possible before getting into doing a buyer’s journey. Estimate Direct as best as possible Even with the best tracking in the world, some traffic is still going to come in as direct. With the rest of your data on sources, you can extrapolate where your direct traffic came from.
How does marketing influence consideration?
For our purposes in this article, we’re going to assume that your main impact is done through the use of the website. Some marketing consultants have started using Google Analytics to map out the most common web page journeys. If you have a very linear, start to finish path, you may be able to easily document it out using the “Users Flow” report in Google Analytics. In my experience, most companies will not get a ton of insight from this report because high-quality prospects make up a small portion of website traffic and they get drowned out by garbage traffic. Instead, I prefer to pull a representative group of real prospect journeys. It’s best if these are recorded in a marketing automation platform and are non-anonymous so we can easily verify lead quality. It’s not very glamorous, but hand-analyzing the prospect journey of 8-12 high-quality prospects can frequently give you a very solid understanding of how the website and marketing content influences those prospects during the consideration phase. At the end of this phase, you should have documented what marketing content affects the majority of buyers along with a couple of example journeys of real customers to serve as proof. For us, we see a lot of users going to our capabilities page. No surprise there, we should make sure our capabilities page is awesome!
How does marketing send the prospect to sales?
It’s not enough to just filter by conversions. Leads sent to sales have to be properly filtered for quality if you want to get correct insights from your marketing data. This is where a marketing automation platform comes in handy because you can filter users in many more ways than you can in other platforms but still integrate with other analytics platforms if you wish. Filter by geographic area, filter out test form fills,
Conclusion: Assertions in a buyer’s journey should be supported with data
Without supporting statistics, assertions in a buyer’s journey are just left flapping in the wind and they won’t be able to withstand a simple challenge by the dreaded HiPPO (highest paid person’s opinion). Providing User Journeys to clients is a high value, high-cost service, and your clients deserve them to be backed up with real data.