The freelance client context I need at the start of every call

Before every freelance client call I check five things: the last decision they made on the call before, the deadline they're nervous about, who actually approves work, the budget ceiling they implied but never put in writing, and the one tool they refuse to switch from. Ninety seconds. Every kickoff and every check-in.

Tuesday, 9:54 AM. You're three minutes into the second coffee, six minutes from the standing call with the SaaS client in Austin. Chrome is open, Slack is open, Notion is open, the Gmail thread with the signed scope sits in another tab. The Loom the marketing lead recorded last Thursday, where she paused and said, 'actually, can we make the hero copy more direct,' is somewhere. You search 'hero copy' in Slack. Nothing. You search Gmail. Eighteen threads, none of them right. The Notion comment about the legal team approving claims is on a page you can't remember the title of. At 9:59 you join the Zoom with a blank notebook and the small, familiar dread of being one beat behind.

The five things I look up before any client call

The list is short on purpose. Long checklists become their own kind of clutter.

The last decision they made. Not the decision you remember making. The decision they made out loud on the last call, ideally in their own words. If a copywriter client said 'stop using the word innovative, it makes us sound like every other Series B,' that's the line you want in front of you. Not paraphrased. Verbatim, if you have it.

The deadline they're nervous about. Most clients have one date they care about more than the others. A product launch. A board meeting. A 1099 cycle. A trade show in March. Knowing which date is the load-bearing one changes how you push back on scope.

Who actually approves. The person on the call is not always the person who signs off. The marketing manager runs the standup; the VP has final say on hero copy; the legal team owns claims. Misread the org chart and you'll redo work twice.

The budget ceiling they implied. Sometimes it's a number ('we have about 8k left this quarter'). Sometimes it's a phrase ('we don't want to go through procurement again'). Either way, it's a constraint they won't repeat unprompted.

The one tool they refuse to switch from. They use Figma, not Sketch. HubSpot, not Salesforce. They write copy in Google Docs because legal redlines there. Push them off it and the project stalls.

Two of my retainer clients renew quarterly; getting these five right on the first call of a quarter is what separates a renewal from a 30-day kill clause.

Why retrieval, not storage, is the actual problem

You probably saved all of this. The Loom link is in a Slack DM. The brand voice doc is starred in your bookmarks bar. The budget conversation is in a Gmail thread. The wireframe screenshot is in a folder called 'client-name-2026' on your desktop. Saving wasn't the failure. Finding, at 9:57 AM on a Tuesday, was the failure.

This is the part most productivity advice gets backward. The pitch is always: better folders, cleaner tagging, a more disciplined inbox-zero ritual. None of that helps when the question you're trying to answer is not 'where did I file this' but 'what did she actually say about the hero copy.' You don't remember the folder. You remember the moment.

Across the freelance subreddits and the browser bookmark graveyard most of us are carrying around, the pattern is the same: people save more than they retrieve, and the ratio gets worse every quarter. By the end of 2025 I had something like 4,200 bookmarks across two profiles in Chrome. I retrieved maybe forty of them on purpose.

How I save context during a call without organizing anything

The trick that's worked for me, for about eight months now, is to stop trying to file things during the call. I used to keep a Notion page open in a second window and try to take structured notes in real time. Headings. Tags. Backlinks. By month two of any engagement the page was 4,000 words of out-of-order bullet points and I dreaded opening it.

Now I do the opposite. Three things, in any form, no taxonomy:

1. A rolling text file (literally just a .txt on the desktop, named after the client) where I dump quotes as I hear them.
2. A Chrome extension that grabs the Loom link, the Slack permalink, or the Notion comment URL with one click.
3. A Telegram thread to myself where I forward email screenshots and voice memo myself if I'm walking after a call.

The point is not that this stack is special. The point is that none of it requires me to decide where something goes. I'll figure out 'where' when I need to find it, not when I'm saving it. Save it, forget it, ask for it later.

This is the same pattern people are starting to call recall-first second brain thinking. Saving is cheap. Retrieving is the bottleneck. Treating those as separate problems, instead of solving them with one Notion database, changed my prep time from 12 to 14 minutes per call down to about three.

What the existing tools actually do well (and where they leave gaps)

Notion is real software with real depth. If you have the patience to build a client CRM in it, with linked databases for projects, calls, and deliverables, it will hold up for years. The catch is the building. I've watched freelancers spend a Sunday afternoon, twice a year, rebuilding their Notion workspace because the old structure stopped fitting how they actually work.

Obsidian and Logseq solve a different problem. They're good for knowledge you want to link and revisit on your own schedule. Less good for the question 'what did this specific client decide on May 14.' That's not what graph-based notes are built for.

ChatGPT Memory and Claude Projects can hold context, but they're scoped to chat. If your prep happens across Loom, Slack, Gmail, and a Google Doc, you're still the one ferrying material into the chat window. That part hasn't been solved by any of the LLM memory features as of mid-2026.

What's actually missing, for the freelance kickoff scene specifically, is something between a notes app and an inbox. You want to save fast (no folders, no tags, no organizing), and you want to retrieve by meaning, not keyword. That gap is what search notes by meaning, not keyword describes. It's also where dEssence sits: memory you don't have to maintain, accessed through a Chrome extension, Telegram bot, or the web app at dessence.ai as three co-equal save surfaces.

Honest about dEssence

I named Notion, Obsidian, Logseq, ChatGPT Memory, and Claude Projects above, so it's only fair to say what dEssence isn't yet. It's in public beta. There's no native iOS or Android app, which means if you're walking out of a coffee shop and want to capture something fast, you're using the Telegram bot, not a dedicated mobile app. The free tier caps archive size, which is fine for a couple of clients but worth knowing if you're juggling a portfolio of ten. There's no team or shared collection feature, so handing context off to a subcontractor still means copy-paste. The paid tier isn't finalized.

Frequently asked questions

Q: What's the minimum I should review before a freelance client call?

The five things I open this article with: last decision, anxious deadline, real approver, implied budget, sticky tool. If you only have ninety seconds, those five give you about 80% of what you'll need.

Q: Where do I save client context if I don't want to maintain folders?

Anywhere fast and durable. A .txt file on the desktop, a Chrome extension that captures URLs in one click, a Telegram thread to yourself for voice notes. The principle is no folders, no tags, no organizing at the save step. Retrieval is a separate problem.

Q: How do I find a quote a client said weeks ago without remembering the keywords?

This is what meaning-based search is for. You ask in your own words, the way you'd describe the moment to a coworker ('what did she say about the hero copy on the Loom last week'), and the system surfaces matching context. Keyword search fails here because clients rarely use your search terms.

Q: Is one client-context system enough, or do I need a different setup per client?

One is enough if retrieval is fast. The bottleneck for most freelancers isn't mixing clients; it's finding the right context for the right client when the call starts. A single archive plus per-client tags at retrieval time tends to beat per-client workspaces.

If you want to try the recall-first approach without committing to building it yourself, dEssence is free during beta, no card. It's the Chrome extension, Telegram bot, or the web app at dessence.ai, with one retrieval layer over all of them. It's beta, the mobile apps aren't here yet, and the free tier caps archive size, so it's worth knowing the limits before you migrate years of context into it.