Coaching that starts from the rep’s own call.
Every conversation begins with the moment it went wrong, in their own recording — not with a general remark about listening more.
- Fix
- Discovery ends before the budget question
- Say
- “Who else signs this off, and what do they need to see?”
- Because
- “I’ll need to run it past the team first.”31:08
The one-to-one is on Friday, the call was on Monday.
By then nobody remembers the wording, and the talk turns into advice in general terms. Advice in general terms is the kind people agree with and never apply.
Coaching happens when there is a gap in the calendar
You know who needs an hour with you. You also know that hour has to come out of a week that is already full, and that by the time it happens the call you wanted to discuss is three weeks old and nobody remembers it.
Coaching from memory is coaching about the wrong thing.
What changes
One thing at a time
The gap that costs the most, not the seven you noticed. Seven changes at once means none of them.
The exact sentence to use
Not “handle the objection better” — the words, and a colleague’s recording where they worked.
Next week shows whether it stuck
The same measure a week later: did that moment change, on real calls.
How it works on your conversations
Every screen below is the product itself, running. Try one.
Decide who gets the hour before the week starts
The team table sorts itself worst first, so the person who needs your time is already at the top. Each row carries the movement since last week, which separates the rep who is struggling from the rep who is improving slowly and needs to be left alone.Who to coach first, decided by the calls rather than by who spoke up.The week, by person
Sorted worst first. The list already knows who needs the hour.
Point at a row to see what is behind the number
Arrive with three things, taken from their own calls
Each rep has three growth areas for the week, written from their own recordings, with what to keep alongside what to change. Three, because a week has room for three. Strengths stay in the plan, so the coaching stops removing the habits that already work.The plan is written before Monday, from calls that happened on Friday.Simon Reed
Week of July 27 – August 2 · 9 calls
Skill map
Eight directions, scored from this week calls
Point at a direction to see what is inside it
What to fix first
Ranked by how many conversations it cost
This week
Three, because a week has room for three
Agree the next step before hanging up
Seven of nine calls ended on "I will follow up". None had a date.
Before we hang up: what date works for the follow-up, and who else should be on it?
Answer the rollout worry instead of moving past it
Raised in five calls, answered in none of them.
What made it not stick last time, the tool or the habit around it?
Ask what the number is being compared against
Four calls went to terms before anyone said what the budget was.
Above what you planned. What were you working from?
Give them the sentence to use
A task carries the moment it came from, why it cost the deal, and the exact wording to use instead, ready to read before the next call. “Handle the objection better” is the kind of advice people agree with and never apply.The moment it went wrong, and the words to use the next time it happens.Agree the next step before hanging up
31. Next step agreed out loud
→ Where it comes from: the coaching plan for July 27 – August 2
Why exactly this
Calls that end with a named date reach the next stage far more often than calls that end on “I’ll send something over”. Across your team this one item is worth +14 points on getting to the next step.
How it goes now
The call ends on “I’ll send you something and we’ll be in touch”. Nobody names a day. The buyer agrees, because agreeing to nothing is easy, and the deal quietly goes cold.
How it should go
Before hanging up the rep names a day, a time and who calls whom, then waits for the buyer to say it back. The step exists only once both people have said it out loud.
How to do it
- 1Two minutes before the end, stop presenting and say you want to agree the next step.
- 2Offer two specific times instead of asking when it would be convenient.
- 3Say who calls whom, so nobody waits for the other side.
- 4Ask the buyer to repeat the date back, and write it down while still on the call.
Ready-made phrases · Copy before the next call
“Before we finish, shall we put the next conversation in the calendar? I have Thursday at 11 or Friday after 3. Which works better for you?”
“So that this does not get lost: I call you on Thursday at 11 on this number. Does that work?”
“Let me repeat what we agreed, and correct me if I got it wrong: on Thursday I send the breakdown, on Friday we go through it together.”
Verify it changed on a later call
Tasks move through columns like work. The column that matters is verification: where a manager confirms the change actually showed up in a conversation that happened afterwards. Without it, coaching is a conversation you had, rather than a thing that moved.Assigned on Monday, checked against a real call the week after.Tasks this week
Written from the calls, assigned to a person, checked against the next conversation.
Press a status to see what it actually means.
- Agree the next step before hanging upNathan GreyTeam 1highReturned
- Ask what the number is compared againstAdam ReyesTeam 4highUnder review
- Name the objection back before answering itAlla KrylovaTeam 1highIn progress
- Summarize the need before presentingElise NavarroTeam 5mediumConfirmed
- Bring the second decision maker into the callMarcus WebbTeam 4mediumNew
- Stop pitching before the buyer finishesJenna SloanTeam 5mediumNew
- Check the request before quoting anythingIsaac LaneTeam 1lowNew
How it lands on the week you already have
- The one-to-one
- Opens on the rep’s own recording at the moment it went wrong, so the half hour goes to fixing rather than to reconstructing.
- Managers with ten reports
- The plan is written before the meeting. A manager reviews and adjusts it instead of building it from nothing.
- Remote teams
- Coaching stops depending on who sits near you or who asks for help. The calls decide.
- One change at a time
- Seven improvements assigned at once means none of them. The gap that costs most goes first, and the rest waits.
- Proof for the rep
- A colleague’s recording where the same moment went well, so the wording arrives with evidence rather than as an instruction.
What happens, and when
Week one
Each rep gets three things to work on, drawn from their own recordings.
Month one
Coaching conversations start from a recording both people have already read.
Quarter one
You can see which coaching moved a number and which did not.
Where to start
Each product is bought on its own. Start with the one that hurts.
Spellit Actions
Start with it!Spellit Academy
Spellit Client Truth
Spellit Notes
See the demo now. Then run it on yours.
A live demo of our products and a real conversation
about the growth problem you need to solve.
FAQ
What does a Spellit coaching plan contain?
A Spellit Actions coaching plan gives each rep one to three tasks for the week, built from that rep's own calls. Alongside the tasks the rep sees three strengths and three growth areas, so the plan protects habits that already work.
Each task opens into a short lesson. It says why this skill was picked, how the rep does it now and how it should sound, and what steps to take. Then come up to three ready phrases, a colleague's call where it went well and a moment from the rep's own call where it went wrong. The lesson ends with what success looks like, the expected result and a suggested session format and length. Phrases copy with one button. The plan also lists the rep's open deals that have no next step.
How does a manager prepare a one-to-one with Spellit?
With Spellit Actions a manager prepares a one-to-one by opening the rep's row on the team board, where the talking points are already laid out. The board sorts people worst first by success score and shows each metric's movement since last week.
Expanding a row shows what to fix first, with the rep's miss rate and the name of a teammate who gets that item right in at least 80% of calls. It also shows the week's tasks and how many of the rep's deals can still be saved. One more click opens the personal review: the weekly report for that one person, a skills map by funnel stage and the week's tasks. Any number opens into the calls, quotes and timestamps behind it, so the meeting starts from the recording.
How does Spellit decide which skill a rep works on first?
Spellit Actions picks the skill to coach first by combining how much it matters for reaching a next step with how often this rep misses it. Impact is measured on your own project's calls, as the difference in how often a conversation reaches a next step when the skill was used versus when it was skipped. The miss rate is counted on the rep's own calls.
Two filters keep the ranking honest. Items missed too rarely to draw a conclusion are left out. So are items that describe the outcome itself, such as the next step or the sale, because their link to success is circular. The other gaps wait their turn, since seven changes assigned at once tend to produce none. The ranking is a calculation, and the AI only writes the lesson around it.
How do we see that a rep actually improved?
Spellit Actions shows improvement by checking the same skill on the rep's later calls, so progress is measured in real conversations. The skills map on the rep's page shows how well they handle each item and group, with a trend arrow, and every figure opens into the calls behind it. On the team board, each metric beside the rep's name carries its week-over-week change, colored by whether that direction is good for the metric.
Tasks add a human check. The rep marks a task done, and the manager confirms it after seeing the change on a call that happened later, or returns it with a reason. Every status change and comment goes into a log that cannot be edited afterwards.
Is the coaching advice written by AI or chosen by rules?
In Spellit Actions a rule chooses the coaching tasks, and the AI writes the words around them. The one to three tasks come from the skills map and the impact ranking, which are calculations on your calls. The AI then writes the lesson from real material: the rep's own quote, a colleague's call where the same moment went well, and phrases to try.
Every recommendation is labeled as AI-generated support, and decisions about people stay with the manager. Because each lesson rests on quotes from actual calls, a rep can open the source and check it before agreeing. Scores and texts are produced once a week ahead of time, so the screen reads a finished result. Nobody waits for a model to answer while the meeting is running.
Can a rep push back on coaching feedback?
Yes, in Spellit Actions a rep can challenge both the score and the task built on it. If the score behind a task looks wrong, the rep files an appeal on that call and rubric item with a written justification. A person then decides and has to explain the decision in writing.
On the task itself, the rep and the manager can talk in a comment thread under the lesson. The manager can send a task back with a reason or dismiss it if it does not fit the person. Each action is recorded with who took it. Spellit labels its coaching as support for the manager rather than a verdict on the rep, and the final word on any score belongs to a person. That matters, because people act on coaching they believe.