Go-to-market operations: remember what happened and what happens next

You've got a reply waiting on LinkedIn, a customer question buried in email, and a follow-up you remember agreeing to send. The date may be in a notebook. It may also be in the special brain-based filing system known as “I’ll remember that.”

This can work while you have three conversations happening at once, but it gets expensive when you forget to reply to somebody who was ready to pay. Therefore, early go-to-market operations should give you a clean and clear place to see who you are speaking to, what happened and what needs to happen next.

Try not to get bogged down into over-complex tools on the market that'll take you longer to set up than it will to respond to those outstanding emails. A spreadsheet, Notion table or lightweight customer relationship management systemCustomer relationship management software. A CRM stores information about people, companies, conversations and sales opportunities. The software only helps when its records reflect the current situation. (CRM) like Folk can all do this; the software matters much less than the info inside it being up-to-date and easy to access

Start with the smallest place

Your first system doesn’t need automated scoring, twelve dashboards or a field for every thought you've had since 1997. It needs to hold the few details that change what you do next.

Whichever option you choose becomes your system of recordThe place treated as the current answer for a specific type of information. In this chapter, it is the table or CRM you check for customer conversations and next actions., meaning the place you check when you need the current answer. If you still trust your inbox or memory more, the record is either out of date or asking you to maintain too much.

Decide who gets a row

A person belongs in the record when they enter a real conversation, show clear buying interest or become a customer. If you sell to businesses, the company and its buying process may need their own details, too.

A website visitor doesn’t become a sales opportunityA person or company considering a purchase. Interest should be supported by customer behavior such as asking for details, discussing price or involving another decision-maker. because your analytics tool noticed them. For a self-serve consumer product, the people worth logging individually may be interview participants, customers who ask substantial questions or somebody who messages before buying. Anonymous visits and product behavior belong in your analytics; we’ll cover those in Measuring what matters. Fifty website visits and five people who asked about price are different facts.

Keep seven details about each conversation

You can add fields later; these seven are enough to tell you what has happened and what you should do next.

For a considered B2B purchase, a few more details may change how you handle the sale.

Record the original sourceThe first known route through which somebody found you, such as a referral, post, search result or cold email. “Unknown” is a valid source when the information is unavailable. when you know it, such as a LinkedIn post, referral or cold email, because this helps you to double down on the channels that are most productive for you. You can also record what prompted the latest action: somebody may first discover you on Reddit, join your email list and eventually book a call after a customer story. “Unknown” is still more accurate than awarding the sale to whichever channel you checked most recently.

Use states that describe what happened

A sales stage is the label attached to the current point in a buying conversation, but it only helps when you can describe the customer action that puts the person there.

Your buying process may need different words. The way a person enters this process matters more than the label; two records in “considering” should represent roughly the same stage.

“StalledA conversation with no agreed next step, or no customer response after the relevant follow-up. Stalled records can be reopened later without appearing active in the meantime.” deserves its own state because an old conversation shouldn’t remain active forever, and you want to record that the person was once interested so you can re-activate the relationship at a later date.

Give every active conversation a dated next action

An active record needs a specific action, the person responsible and a date. “Follow up” leaves this too ambiguous.

The next actionA specific action assigned to a person and date. It may belong to you or the customer, although you still need a date to review anything the customer agreed to do. can belong to you or the customer. If the customer has agreed to do something, the record should still say when you will review it. Otherwise “waiting for reply” can become a permanent state of anxiety. An automated email doesn’t update the date of the last meaningful interactionA reply, call or substantive exchange with a customer. An automated message you sent is activity; it does not show that the customer responded. — it instead records something you sent. A reply, call or substantive exchange is the meaningful point to record.

Separate your activity from customer response

Posting, emailing and attending an event tell you what you did; replies, sign-ups, trials and payments represent your customers’ actions following. Keeping both in the same short activitySomething you or your business did, such as publishing a post, sending an email or attending an event. Activity becomes commercial evidence when a customer responds through an action such as replying, registering, trying or paying. log makes the relationship between the two — the cause and effect — visible.

Time or spend deserves a column when it affects your decision. You don’t need a conversion dashboard for three replies, and a percentage built from a tiny sample won't tell you much. The measurement chapter will handle rates, benchmarks and attributionThe method used to connect a customer action with the channel or activity that contributed to it. Early records can preserve the original source and latest prompt without attempting to assign a precise percentage of credit. in detail.

Record decisions before you forget why you made them

The decision logA short record of a commercial decision, the evidence available when it was made and the condition that would cause it to be reviewed. can be extremely short. It needs the decision, the why, and the condition that would make you review it.

This gives an early decision a fair assessment without turning it into something that might not be true forever.

Clean the record once a week

A good rule of thumb is a 15- to 20-minute weekly clean-up for an early system. This doesn't have to be manual - rope in an AI agent to help you - and treat it as a starting point based on a small number of active conversations. The time is enough to:

• add missing next actions and dates;
• move inactive conversations to stalled or closed;
• correct missing sources;
• record decisions made during the week; and
• check what is due in the next seven days.

If the clean-up regularly takes much longer, remove fields before adding software. Your cash, available time and delivery capacity should also change how you work. A list of 40 follow-ups isn't viable if you can only complete eight, and six new customers aren’t good news if you can only support two.

Automate after the manual process works

Automation helps when the same repetitive actions keep taking your time. Before handing it over, you should be able to answer five questions.

AI can prepare admin work for your review, while customer communication and material changes remain yours to rubber-stamp.

The model can’t know that a polite “interesting” means a $20,000 opportunity, no matter how enthusiastically it fills the field.

Three conversations after one week

Suppose you sell RelayReport, a fictional reporting tool for small marketing agencies. Three people enter your record through different routes.

This represents one progressing conversation, one possible later opportunity and one clear no. Calling all three “open” would make the pipeline look nicer while telling you very little.

If you sell a self-serve consumer product, the individual record will probably be smaller. You might record research participants or design partners, substantial support conversations and people who ask a buying question; your product analytics will count anonymous visits, sign-ups and repeat use.

Prepare your GTM operations record

Use the GTM operations record with any customer notes, channel activity and decisions you already have.

I am preparing my GTM operations record. Use only the material I supply.

Don't contact anybody, send messages, connect accounts, change a live CRM or
invent missing customer information.

1. Recommend one place for my record based on the number of conversations I
   handle, the tools I already update and whether anybody else needs access.
2. Build my people and opportunities table. Use these starting fields: person,
   company where relevant, source, current state, last meaningful interaction,
   next action and date, and the main question or objection.
3. Add likely value, expected decision date, decision-maker status and budget
   status only where my B2B sales process needs them.
4. Classify each record as new, conversation started, considering, customer,
   stalled or closed. Cite the supplied customer action behind the state. Mark
   any state that can't be supported as unknown.
5. Create an activity log that separates what I did from what customers did.
6. Create a decision log with the evidence, date and review condition for each
   supplied decision.
7. Create a seven-day view of my dated next actions. Don't invent follow-ups
   where none have been agreed.
8. Prepare a weekly clean-up checklist and draft AI permissions. Keep all
   external communication and material record changes subject to my approval.

Return standard Markdown in my first person. Separate supplied facts,
assumptions and unknowns. Stop with the prepared record for my review.

The record is prepared when the three tables exist, each active conversation has a dated next action or a marked unknown, and the seven-day view shows what you have committed to do. The next chapter covers how to keep in touch when somebody is interested and still not ready to buy.