In this guide, I’ll share a framework for engaging midmarket and enterprise accounts during the dark times of quasi-personalized AI-generated clutter and sellers struggling to grab the eyeballs of decision makers.

This approach rests on the 15+ years track record in helping enterprise teams from the US to Mongolia in 30+ niches. So even if you work in a non-SASS company and, for example, sell software development or industrial equipment, this framework will get you covered.

A few housekeeping items before we begin

• The framework is tailored for upper-midmarket and enterprise accounts, as it requires severe time investment towards every target account
• Thus, if you are working with SMB clients, proceed cautiously, as the math probably won’t add up.
• Very little will be said about the tech stack (other than basic LLMs and a little bit of Clay for contact sourcing). I firmly believe that the right approach and clear processes are much more important than the bells and whistles of a new outreach platform. Also, there’s no shortage of sales tech advice out there, unlike the battle-born enterprise selling frameworks from non-SaaS domains.
• If you liked what you read and saw, please share the link to this page with your fellow enterprise sellers and marketers.

Now let’s set the mood with a 22-second video (old meme, but still pure gold) … and dive in.

Four steps of engaging an enterprise account with cold outreach

These are the key building blocks that we’ll cover in this framework.

• Assessment — a set of parameters that allows you to rate the potential of the target account and prioritize your efforts.
• Preliminary account research (PAR) — thorough analysis of various data sources tied to the target. The “Artifact” of this stage is a document with dozens of intel nuggets that can be used as icebreakers.
• Value hypothesis — assembling the intel nuggets into highly personalized account-specific “takes” that will cut through the ai-generated outreach clutter and get the attention of the target stakeholders.
• Outreach cadences — spoon-feeding the value hypotheses to various stakeholders inside the target account with personalized cadences.

Let’s talk about every step in exhaustive detail.

1. Assessment

If you send cookie-cutter cadences with minor personalization, one or two “bad apples” in the account list is not painful. But in the situation when prep. work for every target account takes (spoiler alert!) 30+ hours of the team, the mistakes in account selection can be really painful and demotivating.

This is where the assessment model comes into play.

Blocks of the assessment model

Parameter selection and standardization

You will score target accounts based on the set of criteria that’s unique to your business.

Few things to consider when creating this list:

• Please make sure that the parameter comes from an easily attainable data source.
• Headcount growth in the department is something that can be easily found.
• On the other hand, severe internal feuds inside the same department are the data point that’s invisible to the beholder’s eye outside the company (although it would be great to know this stuff before reaching out)
• Parameters should reflect the potential value of the account.
• Treat them as ICP criteria on steroids.
• Also, this is a good place to introduce disqualification criteria (i.e. “no outsourcing policy”) that instantly remove the account from the tier-1 list
• Parameters are likely to be diverse (budgets, so you need a unification scale).
• It is better to start with a min-mid-max scale (I would recommend going with 1-3-5).
• For "yes or no" parameters, use 1 or 5.

Weight Distribution

The importance of parameters in the assessment model will vary greatly. Therefore, we need a system that allows us to evaluate the contribution of each data point to the final score.

My go-to tool here is a good old checkboard method

• You compare each contestant against each other and choose the most important parameter in each pair (without the possibility of a tie).
• The winner gets one point, and the loser gets zero.
• After comparison, the sum of each participant's points is divided by the total number of points played.
• The final value will be the "weight" of the parameter (screenshot with the example)

Assessment

Now it’s time to run every account through the model and select crème de la crème.

Few issues you can stumble upon

• If the spread between min and max is ≤10%, you probably went with the obvious parameters and didn’t dig deep enough. So you would need to return to step 1 and think of additional data points.
• If a parameter requires too much time to extract and its weight is less than 15% (or 0.15) then it’s better to cross it off and recalibrate the model with the algorithm from step 2.

Account Assessment: How to leverage AI

• Drafting ideas for assessment parameters.
• Parsing data points using AI agents (e.g., visit the website to check if the company has clients from major retail).
• Parsing data points from LinkedIn / Apollo / D&B (company size, job openings, development offices, etc.) is more convenient when done through Clay.

2. Preliminary Account Research

At this stage, we need to gain enough account intel to start highly personalized convos with various stakeholders across the account.

It’s easy to get carried away with a variety of data sources, so I strongly recommend you use the “top-down” framework and slowly move from generic intel sources (10 Ks, investor calls, etc.) to contact-specific digital traces (interviews, LI posts, etc.).

Here’s a reference table with the typical intel sources.

Five steps of the preliminary account research

Step 1. Grasp the big picture of what’s going on in the account's industry

You need to feel the zeitgeist: what are the major market shifts, what the leaders are doing etc. Then, understand how your target account is coping with those changes.

What to look for

• What’s going on in the industry (trends, lay-offs, growth, stagnation, geopolitical stuff, etc.)
• How the client’s business works (product line, target industries, etc)

Where to look

• Industry analyst reports
• 10K/10Q reports of the target account
• Earnings Calls (don’t forget the Q&A section, sometimes it can provide crazy insights)
• • Check out my short presentation on extracting data from EC

Step 2. Understand the key initiatives of the account

It's extremely hard to land a large deal if it's not somehow attached to a strategic initiative.

What to look for

• Top priorities inside the account
• “Star Products” that are getting major investments

Where to look

• Investor Day Presentations
• Here's a short post with examples of data that can be extracted
• Leadership Interviews
• Initiatives with wide PR Coverage
• Recent C-level hires

Account Research: how to leverage AI

This is the part where AI tools really shine.

With proper prompting techniques, you can analyze incredible volumes of various data sources (websites, PDFs, YouTube videos, etc.) for a mere fraction of the time compared to the age of manual labor.

My go-to tool at this stage is NotebookLM by Google:

A word of warning — although there are quite a few platforms that claim to conduct a “done for you” account research, their output is usually severely worse compared to NotebookLM or Chat-GPT.

Here’s a 12-minute video that proves my point

Step 3. Deep-dive into the initiatives that can be tied to your offer

Now let's take a closer look at the important projects inside the account that could benefit from your offerings

What to look for

• What initiatives can we tag along?
• What departments are responsible for these initiatives?
• What’s the “connective tissue” between our prop and the client’s initiative?

Where to look

• Job posting and overall hiring dynamics
• Niche podcasts and interviews with C-1 level managers
• Product/customer webinars
• Materials from tech days and user conferences

Step 4. Draft the account map

Now let's get a glance at the potential members of the decision-making unit (DMU).

What to look for

• What can the chain of command look like?
• Who can be involved in the buying process?
• Shout-out to Nate Nasralla and his brilliant article on account mapping (with a small comment from yours truly)
• What department should we reach out to first?

Where to look

• LinkedIn SalesNav
• Zoominfo/Cognism/Similar data provider
• Case Studies by other vendors selling to this account
• Job postings (both current and archived)

Step 5. Research the potential stakeholders

Let's uncover some insights that will help engage with the buying group members (I prefer to start low and get some additional intel before reaching out to the C-Suite).

What to look for

• Relevant KPIs
• Career trajectory
• Contact data
• Personal context

Where to look

• "Casual" social profiles (FB, Instagram, TikTok, Twitter)
• LinkedIn profile
• Zoominfo, Cognism, or similar data providers
• “Creepy” open-source intelligence (OSINT) tools and some social engineering.
• Here's my infamous "$4M ARR with a Blue Jersey story" to give you an example of this approach

Account research results

The main artifact of the PAR stage is a spreadsheet with intel nuggets, that are described in the following fashion:

1. Hypothesis a.k.a. insight — what account-related event can be used as an outreach trigger
2. The reason why — what are the triggers that make us think the offer can be relevant
3. Reference — link to the source that contains the intel nugget.

Here’s how the final doc can look like:

3. Account-specific value hypotheses

In the previous steps, you gathered a lot of scattered insights. Now it’s time to assemble them into super-personalized takes for different business units inside the target account.

In an ultimate must-read, A Practitioner's Guide to Account-Based Marketing, Bev Burgess shared a very elegant visualization of the difference between segment-based and account-based value props:

To create a value hypothesis, I would recommend using a six-step framework (again, taken almost verbatim from Bev’s wonderful book)

Six-step framework for creating a value hypothesis

Result - 5-7 value hypotheses per target account

Here’s an example of what a structured value hypothesis can look like (please keep in mind that during this stage you need to draft several of those hypotheses for different business units as it will maximize your chances of hitting the target).

4. Outreach cadences

Stage Objective: blend the intel nuggets into super-personalized messages tailored for every stakeholder inside the account

Building blocks of the first message

Here’s the typical structure of a first message created and made famous by the Lavender Crew.

1. Trigger based on account intel nugget
2. Assumption based on the trigger
3. Elevator pitch of the value hypothesis
4. Soft CTA (a.k.a. never ask for the call in the first message)

A well-executed research should allow you to easily craft a unique "take" for each potential stakeholder in the target account.

In other words, we are not trying to create an “ultimate outreach” that will blow Mr. Big Decision Maker’s mind. Instead, we launch dozens of highly personalized cadences across the account.

I often call this approach “Death by a thousand cuts” (凌遲). It may sound harsh, but that’s exactly what we are doing here.

Crafting the cadence

Given the intense (almost insane) amount of “targets” inside one account, I would recommend going rather easy with the cadences themselves — stick with 7-8 touches both in email and LI across 10 days.

A few random thoughts:

• The first message will generate at least 50% of all the responses
• The best cadence frameworks are useless without a research-driven value hypothesis
• Use the “working bees first” approach – launch cadences to the bottom and then enrich your C-Level outreach with the insights received from the lower levels.

And don’t forget that It’s impossible to create an engaging message without the intel nugget. No amount of generic wordsmithery will engage the C-Level exec. So first you need a solid data-driven hypotheses and then “package it” into digestible cadences.

Epilogue. Labor costs.

Now to the really sad part. All these bespoke, fully personalized cadences will cost ya’.

In last year's build of the framework, I mourned over the fact that the account-based approach to cold outreach might be unusable by most sales teams due to the extreme amount of labor required for research.

But the AI revolution equipped sellers with the tools that can speed up data extraction by at least ten times and thus severely lowered the entry threshold.

However, I still can’t recommend the framework to the folks closing $5K deals, cause you still need to relocate significant brainpower to craft uberpersonalized multi-stakeholder takes for every target account.

So here is a detailed video walkthrough of the spreadsheet (btw, here’s the source file).

And that sums up the lead generation framework that will allow you to engage almost any tier-1 account.

Key takeaways and next steps

Project artifacts

1. Assessment model for selecting key accounts
2. Account research protocol for gathering intel
3. Framework for creating account-based value hypotheses
4. Framework for creating highly personalized outreach cadences at scale based on account research and value hypothesis

Tangible results of implementing the framework

• Increased productivity of key enterprise sellers
• A scalable framework for routinely engaging midmarket and enterprise accounts
• An accurate and data-backed understanding of resources is needed to attract midmarket/enterprise accounts
• A clear vision of how to scale the system and what additional resources to invest (team, tech stack, ad spend)

Next steps

• If you liked what you read, please share this framework with your fellow enterprise sellers and marketers. “Dark Funnel” is my #1 source of referrals😊
• If you have any questions on the framework, DM me on LinkedIn or drop me a line at dan@dangridin.com
• If you need help with finetuning your enterprise outreach to better engage tier-1 accounts, see how we can work together and book a call.

I guide sales and marketing teams to routinely engage tier-1 accounts around the globe and win the biggest enterprise deals faster with less dependency on the “human factor”. Ping me on LinkedIn or drop a line at dan@dangridin.com