From signing to your first scored queue

There is no build phase and no integration project. Implementation is configuration: teaching the platform who you are, who you want, and how hard to push - then reading the first ranked list before anything is sent.

Four things worth having ready

Onboarding moves at the speed of these four. None of them is technical.

A real prospect list
An Excel or CSV export from wherever you source today: Sales Navigator, a conference list, a CRM dump. Real data, not a sample - the scoring is only as informative as the list you test it on.
Someone who owns the ICP
One person who can settle what a good prospect is. The weighting step stalls when nobody has final say.
Named outreach owners
The people who will actually work the queue, with the capacity you expect from each.
A qualification threshold you can defend
70% is the default. Whatever you choose, it should be a number you are willing to hold the team to.

Seven steps, in order

Each phase produces something concrete. Nothing later depends on work you have not finished.

01
Company profile
≈ 1 sessionYou, with us
We record what you sell, the expertise behind it, the projects worth referencing and the skills on your team. Every AI-drafted message is written from this, so it is the one step worth doing slowly.
What you end up with
- Services and positioning captured
- Reference projects the AI may cite
- Team skills and specialisms
02
Ideal Customer Profile
≈ 1 sessionYou, with us
Roles, seniorities, industries, company sizes and geographies you sell into - plus the disqualifiers. If your ICP is really several ICPs, this is where they get separated rather than averaged.
What you end up with
- Target roles and seniorities
- Industries and company sizes
- Geographies and disqualifiers
03
Weighting and threshold
≈ 30 minutesYou
The nine factors ship pre-weighted and total 100. Adjust them to match how you actually qualify, then set the threshold that decides who enters outreach. Both are editable in Settings afterwards, and prospects can be rescored against a new model with no development work.
What you end up with
- Nine factor weights confirmed
- Qualification threshold set
- Score bands reviewed
04
Owners and daily limits
≈ 30 minutesYou
Add the people who will run outreach and choose how prospects are allocated between them. Set the daily limit globally or per owner - unused capacity rolls into the next cycle and prospects keep their rank rather than moving to the back of the queue.
What you end up with
- Outreach owners added
- Allocation method chosen
- Daily limit set
05
Connect your tools
OptionalYour admin
CRM, calendar, enrichment providers and outreach channels sit behind a connector interface. Nothing here is required to start - the platform runs standalone and connectors can be added later without touching the core.
What you end up with
- CRM connected
- Enrichment provider selected
- Calendar linked
06
First import and dry run
≈ 1 dayYou, with us
Upload a real export. The list is validated, normalised, de-duplicated against nothing on the first run, enriched and scored. You read the ranked output before a single message goes anywhere.
What you end up with
- List imported and enriched
- Scores reviewed against your judgement
- Weighting adjusted if it disagrees
07
Go live
Day oneYour team
The queue starts filling to your daily limit. Owners work their list, approve the drafted openers, and positive replies become sales tasks automatically.
What you end up with
- Queue live
- Owners working daily lists
- Replies classified and routed

Durations are indicative. The platform is built - none of these phases is development work.

Weighting, threshold and bands

The nine factors ship pre-weighted and total 100. These are starting values, not fixed ones.

The nine factors

Role / decision authority20 pts
Industry match15 pts
Services-to-needs alignment15 pts
Company size fit10 pts
Geography match10 pts
Technology relevance10 pts
Outsourcing probability10 pts
AI / digital transformation fit5 pts
Contact info completeness5 pts

What each band means

90–100%Excellent / hot prospect
Contacted first, every cycle
80–89%Strong match
Enters the queue behind the 90s
70–79%Potential match
Queued when capacity allows
Below 70%Low priority
Parked - never enters outreach

The threshold defaults to 70%. Change it and every existing prospect can be rescored against the new model - no development work, no re-import.

How prospects reach a person

Pick one at set-up and change it whenever the team changes. The owner stays attached to the prospect for its whole lifecycle, so nothing becomes nobody's job.

Round-robin
Even distribution across every owner, in turn. The simplest choice, and the right one when the team is interchangeable.
Capacity-based
Owners receive prospects in proportion to the capacity you set for them, so a part-time or ramping rep is never over-filled.
Criteria-based
Prospects route by attribute - industry, geography, company size - to the owner who covers that segment.

Four ways implementations go wrong

None of these are platform problems. All four are configuration decisions that are cheap to get right at set-up and expensive to unpick later.

Averaging three ICPs into one
A profile that quietly covers three buyer types produces noisy scores for all of them. Separate them and run each as its own motion.
Setting the threshold too low
A low bar puts weak prospects in front of your most expensive people, which is the exact cost the platform exists to remove.
Raising the daily limit early
The limit is a quality control, not a safety feature. Volume is not the constraint - a relevant message to a well-matched prospect is.
Skipping the dry run
The first import is where you find out whether the weighting matches your judgement. Reading that output is cheaper than un-sending a week of outreach.

What holds at every stage

These are not configuration choices - they are how the platform behaves regardless of how you set it up.

Human-assisted by default
Connection requests, engagement and message sending are performed by your own team from their own accounts. No bots, no browser extensions, no shared credentials.
A complete audit trail
Score computed, owner assigned, message generated, reply classified, task created - every state change is written to the activity log with a timestamp and an actor.
AI you can inspect
The overall score, the factor breakdown, the AI reasoning and the exact input data are all stored. AI output is monitored and reviewable, never treated as settled fact.
Nothing is locked in
Enrichment providers, prospect sources and outreach channels sit behind interfaces. Swap a provider, add email or WhatsApp, plug in a CRM - the core is untouched.

Talk through your set-up Read the product overview

Start here

Bring one export. Leave with a ranked queue.

Tell us how your team sources and qualifies today, and we will show you the scored, ranked, owner-assigned queue that comes out the other side - and what it replaces.

Book a demo See the platform

Not ready to talk? Start the 14-day trial on your own list instead.

14-day free trial
No credit card
A person replies within 24 hours

Questions about getting started

What you need ready, what usually goes wrong in week one, and how to tell it is working.

Want a hand with the first run?

We will sit with you through the first import and the first queue. No charge, and no obligation after it.

Book a 30 minute call All FAQs

How much of this can we do without you?

All of it. The guide is the same sequence we would walk you through, written down. Onboarding help is there if you want it, but nothing here needs an implementation call to unlock.

What do I need ready before I start?

A prospect list as Excel or CSV, a clear sentence about who your best customer is, and one person who will own the queue. The company profile fills itself in from your website.

What is the most common mistake in week one?

Setting the daily limit too high and the threshold too low. Both make the queue longer without making it better. Start narrow, read what goes out, then open it up.

How do I know it is working?

Compare the top of the queue against the prospects you would have picked yourself. When the two agree, the weights are right and you can stop reading every row.