Using AI for your Account Scores
ChatGPT App Script inside
If you’re new to this newsletter then welcome! Thank you to the 1,400 revenue operations professionals who continue to subscribe to this newsletter. You’re the reason I continue to write each and every week on a Go To Market related topic. When I have a template to share paid subscribers will get access. I don’t have all the answers in revenue operations. That’s impossible because RevOps can be uniquely situated for each unique situation. But what I hope you can take away a few guiding principles or tactical snippets which you can use in your day to day. Before jumping into the newsletter, let’s hear from our sponsors that keep most of this newsletter free to readers.
This week’s newsletter is brought to you by BoostUp. The infamous Bermuda triangle of CRM, Spreadsheets and BI scatters data across different islands and complicates decision-making. Centralize your RevOps processes effortlessly and consolidate your field on RevBI. Book an Intro Call with BoostUp
RevOps aren’t system admins. Growblocks is building the 1st RevOps platform, giving you full funnel analytics, continuous monitoring and revenue planning tools - so you can become strategic. Check it out here.
This weeks newsletter is brought to you by Cognism. They give you access to thousands of B2B leads to add fresh data into your CRM, clean stagnant records and fill in any gaps. Enhancing your CRMs data should be simple – and it is with Cognism.
Lately I’ve been working on scoring my accounts. This will help identify which accounts take priority over others. Collaborating with marketing, the account score should allow you to develop your 1-to-1 (outbound+ABM) and your 1-to-many (outbound or inbound + non-ABM) strategies.
Aside from all that, there are several reasons why a RevOps team should develop account scores. These include:
Identifying high-value accounts. Account scores can help identify accounts that are more likely to generate revenue, either now or in the future. This information can be used to prioritize account outreach and engagement, and to focus marketing and sales efforts on the accounts that are most likely to convert.
Prioritizing account outreach and engagement. Account scores can help prioritize account outreach and engagement. By understanding which accounts are the most valuable, the RevOps team can focus their time and resources on the accounts that are most likely to generate results.
Measuring the effectiveness of marketing and sales campaigns. Account scores can be used to measure the effectiveness of marketing and sales campaigns. By tracking the changes in account scores over time, the RevOps team can see which campaigns are having the most impact and make adjustments as needed.
Improving customer relationships. Account scores can help improve customer relationships. By understanding the needs and priorities of each account, the RevOps team can provide more personalized and relevant service. This can lead to increased customer satisfaction and loyalty.
Where can I get started?
Similar to a Lead Score, an Account Score takes into consideration internal and external data points. For a Lead Score it would look something like this:
Internal: contact information (data points include industry, title)
External: behavior (i.e. demo request, page visits) or intent (data points include intent data from platforms such as Cognism)
For the Account it’s going to be similar:
Internal: firmographic information (data points include revenue, industry)
External: behavior (i.e. three contacts are MQLs within a 30 day span, intent data, etc.)
I previously talked about setting up an Account Based Acquisition approach. Read more here.
Other obvious factors that you should consider in an account score include… Also this is more relevant for B2B focused companies.
Company size. The size of the company is a good indicator of its potential revenue. Larger companies are typically more likely to generate more revenue than smaller companies.
Industry. The industry that the company operates in can also be a factor in the account score. Some industries are more profitable than others, and some industries are more likely to adopt new technologies.
Revenue potential. The potential revenue that the company could generate from your product or service is another important factor to consider. This could be based on factors such as the company's size, industry, and growth trajectory.
Level of engagement. The level of engagement that the company has with your product or service is also a good indicator of its potential value. Companies that are more engaged with your product or service are more likely to renew their subscriptions and recommend your product to others.
Competitor analysis. It is also important to consider the company's competitors. Companies that are facing strong competition are more likely to be looking for new solutions, which could make them more receptive to your product or service.
Strategic fit. The strategic fit between your company and the target account is also an important factor to consider. If your product or service is a good fit for the company's needs and goals, then the account is more likely to be valuable to your business.
These are just some of the factors that you should consider in an account score for a B2B SaaS company that sells mostly to software and Fintech companies. The specific factors that you choose to include in your account score will depend on your specific business and goals.
Here’s how I did it before
At one previous startup I set up Account Scores and a Total Addressable Spend metric. The Account Score was used to determine which accounts to prioritize. The Total Addressable Spend was a proxy for the value of that account.
The Account Score I broke down into several weights:
Helpdesk they used: 30%
If they located zd-assets then they found a company that used Zendesk. If they located kustomer then they used Kustomer. This allowed us to quickly qualify in or out the prospect because our product either integrated with that helpdesk or not.
So how would I be able to fill in this data at scale? If that data isn’t available with my enrichment providers then I have a couple of options:
Do the work myself or spread it out across the team. This is an arduous and expensive process.
Shift the work to an offshore team. This is inexpensive but the quality may be suspect.
Build automation to webscrape the information. I ended up using this route since I am fairly handy at writing Python.
Purchase data from webscraping firms such as Builtwith. Btw, I heard that Builtwith is a one person startup. THAT’S CRAZY TOWN.
Using off-the-shelf data from third party providers isn’t enough to compete in my opinion. We have to get creative in how we source information. I believe the next wave of developing ICPs, personas, and account scores will incorporate many of these unconventional tools:
BPOs (offshore resources): after all, we are a global economy
Webscraping: this will eventually move to friendlier interfaces and rely less and less on very technical team members
LLM/AI tools: these tools are VERY interesting. They have the downside of “hallucinating” but when prompted correctly these tools can cite the relevant sources needed to get you 80% of the way there.
The AI/LLM path is intriguing because it’s good enough at a very affordable price point. For example, recently I was tasked with scoring 5,000 accounts that required scouring 10K/Annual Reports and identifying specific types of information. I timed myself. After a couple of hours of doing this I started to move very quickly. I was able to verify 10 accounts per hour! AMAZING!
That means if I don’t tire out then I can complete the task within 500 hours. I’m not going to publish my salary but let’s say for example I made $150K. Using a 52 week, 40 hour work week my hourly rate would roughly be $75. Multiplying out the project timeline then that would only cost my company $37.5K. What a steal.
Compare that with a custom Python and Appscript function I wrote plugging into ChatGPT I was able to score all of the accounts for less than $10. INSANE!
I would strongly consider poking around these tools. Manually verify the information. Then when you get to a place where your prompts increase your precision to 80%-90% trustworthiness, you’ll be able to reduce not only the cost, but speed up the organization’s path to an effective account score.
If you’re a paying subscriber here is the ChatGPT function I used in Google Sheets. If you’re not sure how to do this follow these steps:
Copy+Paste the code below
You’ll need an API key. Go into OpenAI (ChatGPT) to set up your billing and an API key. Paste that back into the App Script editor.
Once you do that you’ll be able to ask a question to ChatGPT by typing in the function =RevOpsImpact_GPT(“question”).
Give it a try and hit me up if you want to chat about incorporating AI tools into your use cases.
ChatGPT code 👇
Whenever you're ready, there are 2 ways I can help you:
1/ If you’re looking to further develop your Revenue Operations knowledge sign up for my courses in partnership with the RevOps Co-Op. → Unleashing ROI course. A ten-week virtual, live instruction RevOps course designed to level up your RevOps Impact (R.O.I.). Lessons from my career scaling from $10M to $100M+. Join 50+ alumni. https://www.revopscoop.com/courses/unleashing-roi → Sales Ops Masterclass. A six lecture virtual, live instruction SalesOps course designed to take your sales operations skills to the next level. https://www.revopscoop.com/courses/sales-operations-masterclass → CS Ops Masterclass. A five lecture virtual, live instruction CS Ops course designed to take your CS operations skills to the next level. https://www.revopscoop.com/courses/cxops-masterclass 2/ Promote your Revenue focused startup to a newsletter with over 900 tenacious revenue leaders. My eventual goal is to shift this to a completely free newsletter for my readers through sponsored ads. Reply to this email if you’re interested in receiving a media kit to learn more.
Keep reading with a 7-day free trial
Subscribe to RevOps Impact Newsletter to keep reading this post and get 7 days of free access to the full post archives.