Workflows Weekly #20 🚀

Hey there!

Here’s the twentieth edition of Workflows Weekly, with a few workflows, tools, and ideas you want to rely on

Let's get into it.

Most website meeting forms only capture name, email, and maybe company.

Everything else - qualification, research, routing - falls on your reps to do manually.

Martin from our team built the Meeting Form Orchestration Playbook to automate it.

Here’s how it works:

Capture form submissions from a tool like Tally or Typeform and route them into Clay.
Enrich missing data and qualify against your ICP and persona criteria with Clay
Sync companies and contacts into HubSpot, then assign the right account owner.
Notify the rep on Slack and trigger outreach through Instantly and HeyReach.

This way, every form submission gets enriched, qualified, routed, and worked the moment it comes in.

Out of all people who land on your website, only a handful end up booking a meeting.

The rest leave without a trace, even though plenty of them are a fit.

That's the problem RB2B solves.

It identifies 80% of the people visiting your site and pushes them straight into your CRM.

Best part about it:

If you are driving traffic to your site, this is the easiest way to see who is showing up.

→ Try RB2B

Dan condensed 2 years of building AI-first ABM systems into a single cheat sheet.

Tech stack, ICP tiering, signal tracking, awareness scoring, metrics, troubleshooting.. the entire framework on one page.

Here's what it covers:

The full tech stack broken down by category (intent, CRM, contact data, sequencers)
ICP tiering from Dream Accounts down to Potential-fit
Awareness scoring from Identified through Selecting
Primary metrics to track whether ABM is actually working
A troubleshooting table for the most common bad results and how to fix them

If ABM is part of your motion, this is the closest thing to a one-page playbook you'll get.

OpenAI just released GPT-5.5.

It's a big upgrade over GPT-5.4, especially for agentic coding and computer use.

Give it a messy, multi-part task and it can plan, use tools, check its work, and keep going until it's done.

Here's what's new:

Handles long, multi-step tasks without losing context or stalling midway
Better at debugging code, navigating unfamiliar codebases, and operating computer interfaces
23% more factually correct than 5.4, with fewer hallucinations
Same speed as 5.4 but uses fewer tokens to finish the same task
State-of-the-art on coding benchmarks at half the cost of competitive frontier models

Available now in ChatGPT, Codex, and the API for Plus, Pro, Business, and Enterprise users.

→ Read more

That’s a wrap for this week.

What part of your content or GTM workflow feels hardest to manage right now?

Hit reply and let us know. We’ll break it down in a future issue.