Data Management: Setting Marketing Ops Foundations for Success

Standardization Above All Else Data is only useful if you can make sense of it. If your data has different formats and values, each across a siloed system of record, you can’t see valuable insight into how your marketing is performing. In order to gain this insight, standardization is critical. Take lead source capture, for example. If you have antiquated iterations with inconsistent spellings or abbreviations floating around, you’ve got some barriers to overcome. How do you make sure everything is standardized, and providing real insights?  A Four-Layer Model For Structuring and Standardizing Leads Marketing automation systems and CRMs are literal to a fault, so keep your list short. This starts from the broadest and goes down in granularity.

• Lead Source – This is the highest level, broadest value. Organic search, paid search, paid social, event, etc.
• Lead Source Detail – One level of granularity past the Lead Source. It might be Google, Bing, Twitter, Facebook, or specific event type.
• Most Recent Action – The offer type they took you up on. Asset download, etc.
• Most Recent Action Detail – The actual asset.

Why do it this way? You can always bundle things up to a higher level report, but you can’t dig deeper if you’re only using broad values.  Consider the next layer — that bottom layer of exactly what someone was doing. When you put tools in place to capture everything from A to Z, you’re able to pull together sliced and diced reports if/when they’re requested later. And when you want to show ROI, you have the data to report on it at any layer. Putting the technical pieces in place for fully structured, standardized data capture begins with a list of your ideal, most comprehensive values.  Tip: Think of your marketing automation platform (MAP) as one person able to function as one hundred. Leverage its power to do as much work as possible.

• Set up workflows to listen for anything that could identify attribution. Look for UTMs being pumped in from URLs, as well as page visits. Also included can be referrer keywords, original search engines, or search terms.
• Stack-rank the workflows in the order you know are the most accurate/likely to be identified, and have the MAP reference the list any time an action is taken, looking for anything that matches. Once it finds a value that identifies attribution, have it stamp that value.
• There will always be some that don’t match any of the rules. For this, Kendall uses what she calls a “Robomaid” campaign, looking for anyone who hasn’t gotten a lead source or lead source detail within 24 hours of the action being taken. This gives the MAP 24 hours to try its best based on the workflows above to identify the correct values, but if it can’t come up with anything, the value gets stamped as “unknown”.

Unknown can have a negative connotation, but it’s okay if it’s carefully and thoughtfully stamped. On a regular basis, go through those unknowns, trying to tweeze apart trends that will help you stamp more of them higher up in the flow. This helps fill the “gray space” that unknowns tend to bring with them. In a sense, this is a version of machine learning: listening and trying to attribute the most accurate or the most well-assumed value. It’s better to have “unknown” data that you have thoughtfully and carefully classified as such, than dead data with nothing attributed to it. Don’t Lose Sight of the ‘Ideal State’ Marketing ops teams must be able to balance their current state with their ideal state, and ensure they’re setting up foundations for success while considering what the business needs will be later on down the line. How do you do both in parallel? Start determining your ideal by speaking directly with your stakeholders in marketing and other departments.

Ask: What would you be able to measure, and what would those data points help you do? The answers you receive are the business definition (aka “plain English” definition) of your ideal state. 

The second part of that question — “what would those data points help you do?” — is critical. When you understand the purpose of the initiative, you do a far better job at discerning whether the data point they’re requesting is in line with the help they actually need. They may actually require a related data point that can be used in a more effective way to trigger a workflow or pivot a report on.

Will they be presenting to C-levels around certain projects? Are they trying to prove ROI? With contextual information, marketing ops can figure out how to best fulfill the ideals — and it may look different from what was originally requested. 

Next comes the reverse-engineering of that business definition. Look at the different fields and data points being captured in your systems to determine what points actually reflect the information your stakeholders want. What comes out of this is the operational definition of your ideal state.  According to Kendall, anticipating what the team may be asked to report on later down the line is crucial to setting up systems for success. Kendall uses Salesforce campaigns to capture anything that requires someone to take an action — whether it be an asset download or demo request. If you’re attaching values to actions consistently and using naming conventions accurately, the reporting possibilities are, “endless in reverse”.

How to Know If What You’re Doing is Working

Marketers typically look at performance metrics to know if their marketing activities are working, but how can marketing ops professionals, whose work is much more behind-the-scenes, know if what they’ve implemented is successful?  It begins with a mindset: Anticipate what you might be asked for down the road. Some of which is, obviously, revenue attribution and the ROI of different channels. When you think of Marketing Ops’ job as being able to report on anything, now or further down the line, you put important practices and tools in place to make that happen. Here are a few Kendall recommends:

Dashboards are your friends

Dashboards can provide the insight you need to determine effectiveness of your marketing operations. Look at them daily, along with your campaign membership and MAP workflows. You get a good sense pretty quickly when you build something what it should look like on a daily basis. If anything looks off, you’ll know right away because you have a clear picture of what it should look like.

Review reports that look at different data points (e.g. lead source) frequently

There will always be a bunch of leads that don’t fit any of the eight to ten values you put in. Use reports to sieve those pebbles — your anomalies — out of the sand. From these, you can find common variations, such as someone using UTMs that are made-up, or reversed, or spelled slightly different. Then, update your reporting structure to roll the anomalies up to their related attribution point — and better yet, inform your team, who can double-check that they’re using the correct values. Kendall uses a bank of “on the fly” Smart Lists in Marketo to do this reporting frequently.

Use a data enrichment tool

Each tool has its strengths and limitations. One may offer real-time appends, while another may append after people are in the system. Clearbit, for example, relies mostly on email domain to retrieve corporate contact information and details, whereas DiscoverOrg can often do the same job using a job title plus company name, without needing an exact name or email address.

Use a platform that shows engagement on the account level

Account engagement is some of your most telling data. Being able to see what accounts as a whole are engaging with your website or other properties can inform how you prioritize and go after certain markets. It’s technically doable with a marketing automation system through custom automations, but ideally, having a platform that does the heavy lifting for you saves everyone time (and money) in the long run.

And finally …

Don’t be reactive; think scalably

In marketing, period, but in marketing operations especially, you’ll often hear from people only when things are exploding. It comes with the territory. It’s all too easy to slide into a reactive, throw-a-Band-Aid-on-so-everyone’s-happy-for-now state, but it always comes back to bite you in the long run.

Some things might take longer, and you might have to explain it to your stakeholders, but ultimately think about solutions that are scalable. Consider options that are going to work long-term and not require you to constantly go in and patch them up. That’s Marketing Ops at it’s best.