What exactly is the Box of Authority?

The Box of Authority is the set of hard coded boundaries, rules, and escalation triggers you define for your AI agent. It ensures the agent only operates within your approved guidelines, so you maintain full ownership and visibility over every conversation.

Is Antares meant to replace my sales team?

No. Antares supports roles by shifting repetitive, time sensitive tasks like immediate lead chasing to the AI, freeing staff to focus on high value human interactions such as tours and relationship building.

How long does it take to get up and running?

Antares follows a structured two phase, 20 day implementation plan: five days of strategy mapping, a 10 day integration and training window, and a final five days for validation and launch.

Can AI agents handle bookings and CRM updates?

Yes. Antares is designed to book tours and trials once a lead is ready and to update records and outcomes within your existing CRM.

How do you ensure the AI sounds like my brand?

During Phase 1, we collect your business data and brand guidelines to tailor the model, then train and test it against real world scenarios to ensure responses match the quality of your top team members.

What if a lead asks a question the AI cannot answer?

Specific escalation rules handle edge cases and sensitive topics. If a scenario falls outside the Box of Authority, the conversation is automatically routed to your human team for a seamless handoff.

Can Antares hallucinate and promise free memberships?

No. Antares uses grounding and only pulls information from your approved knowledge base. If it does not know the answer, it escalates to your team.

Does Antares replace my CRM?

No. Antares integrates with your existing tech stack and updates your CRM so your source of truth remains unchanged.

What if my staff question job security?

Antares is positioned as a leads assistant. It handles the grunt work of after hours follow ups and lead chasing, allowing your staff to focus on high value tours and closings.

What happens if a lead asks a question Antares does not know the answer to?

Antares is technically prohibited from guessing or making up pricing or other information. If a query falls outside its approved knowledge base, it triggers an immediate human handover and flags the conversation for your team.

Will it sound like a clunky chatbot?

Antares uses natural language rather than rigid decision trees and communicates in your brand voice. During validation, we fine tune the personality until it is indistinguishable from a high performing team member.

Is our prospect data secure?

Yes. Antares is enterprise ready and compliant with standard data protection regulations such as GDPR and CCPA. We do not use your proprietary lead data to train public AI models; your data stays within your silo.

What if Antares makes a mistake?

Antares maintains an audit trail for every word it says. If an edge case occurs, the guardrails can be adjusted so the specific mistake does not happen again across your locations.

How much work is required from my IT team?

Minimal. Keepme handles the heavy lifting of integration. Your team provides API access to your CRM and calendar and approves the initial knowledge base. Most setups require less than 5 hours of total human time from internal staff.

Member Growth · AI Era

Membership Growth In The AI Era

A practical operating workbook for fitness operators. Diagnose your funnel, build your business case, and govern your rollout.

Execution Baseline Snapshot

Make lead leakage visible by channel and stage. Complete once, then revisit quarterly.

1.1 Period & Scope

Enter the date range you're drawing your data from. Use a 30–90 day window. Pick a representative period — not your best or worst month.

Period Start Date ?

Period End Date ?

Total Leads (all channels, same period)?

Sites Included ?

1.2 Lead Volume by Channel

Enter leads by channel. % of Total calculates automatically. If a channel is active but you don't have the number, type "Unknown" — that itself is a signal.

Channel	Leads (#)	% of Total
TOTAL	—	100%

1.3 Speed & Coverage

Average first real response time in minutes and response coverage rate per channel. If unknown, leave blank and note it — leakage is present but unmeasured.

Channel	Avg 1st Response (mins)	Response Coverage (%)	Out-of-Hours %

1.4 Booking & Attendance Outcomes

Enter your current baseline rate and your target for each metric. The gap calculates automatically.

Metric	Baseline Rate (%)	Target (%)	Gap

⚡ Funnel Leak Analysis

Based on your inputs, here is where your funnel is losing revenue.

These flags are generated automatically from the numbers you entered above. Green means this stage is performing well. Amber means there is a gap worth closing. Red means this is a significant leak — the highest-return place to focus first.

1.5 Your Diagnosis

Using the Funnel Leak Analysis above, record your conclusions here. What is the single biggest constraint in your funnel right now? Select the category that matches your most critical red flag, name the leak in one sentence, and identify the channel that needs attention first. This diagnosis carries through to your Business Case — the primary constraint and urgent channel you record here inform which lever to pull first in Step 3.

Primary Constraint *

Speed

Coverage

Booking

Show Rate

Recovery

Biggest single leak

Most urgent channel to fix

KPI Dashboard

Weekly performance tracking. Update "This Week" every week. vs Baseline and vs Target calculate automatically.

Weekly review discipline: Start every session with four numbers — Speed to Lead · Response Coverage · Lead-to-Booking · Show Rate. Identify one corrective action. Write it down.

Weekly Action Log

After every weekly review, record one row per issue. Update Status the following week before the meeting starts — not during it.

Week	Issue Observed	Corrective Action	Owner	Status

Business Case

Leads → Bookings → Visits → Members → Revenue. Enter your numbers, see the opportunity.

Step 1 · Baseline Inputs Same period as your Execution Baseline

The fields below have been pre-filled from your Execution Baseline where data is available — leads, and your three funnel rates. Review them to confirm they are correct. Add your marketing spend, average monthly membership revenue, average length of stay, and optionally your gross margin to complete the model.

Total Leads (all channels) ?

Total Marketing Spend ($) ?

Lead-to-Booking Rate (%) ?

Booking Show Rate (%) ?

Visit-to-Member Close Rate (%) ?

Avg Monthly Membership Revenue — ARPM ($) ?

Avg Length of Stay (months) ?

Gross Margin (% — optional) ?

Step 2 · Baseline Funnel Output

What your funnel produces today — calculated automatically from Step 1.

These four numbers are your starting point. They show exactly what your current lead volume, funnel rates, and marketing spend combine to produce right now. Every figure in Steps 4 and 5 is measured as an improvement on this baseline. If the numbers look lower than you expected, that gap is the opportunity.

Booked Visits

—

Attended Visits

—

New Members

—

Current CPA
—
Cost per new member

Step 3 · Improvement Assumptions

This section shows you what your revenue looks like if you improve each conversion rate — even slightly.

The three rates below are the levers. Each one represents a stage in your funnel where leads are currently being lost. Enter a projected improvement for each — the model instantly recalculates how many additional members you would gain, what your new cost per acquisition becomes, and what the revenue impact is over a member's lifetime.

You are not committing to a number. You are asking: "If we fixed our response speed and follow-up consistency, what would that realistically be worth?" The Conservative Preset gives you a defensible starting point: Lead-to-Booking +10pp, Show Rate +5pp, Close Rate unchanged. Even those modest improvements typically produce a compelling result — because the funnel compounds.

How to use this: Look at the Current column — that is your baseline rate pulled from Sheet 1. Enter a higher number in the Improved column. The Change column updates automatically to show the uplift in percentage points, and Steps 4 and 5 instantly recalculate your new member count, CPA, and lifetime revenue.

Conservative Preset (recommended): Lead-to-Booking +10pp · Show Rate +5pp · Close Rate unchanged.
Moderate Preset: Lead-to-Booking +15pp · Show Rate +8pp · Close Rate +3pp. Use only if you have clear evidence of large structural leakage in your baseline.

⚠ The Improved value must be higher than Current. If you enter a lower number the model will show a negative result.

Rate	Current	Change (pp)
Lead-to-Booking Rate ?	—	—
Booking Show Rate ?	—	—
Close Rate ?	—	—

Step 4 · Improved Funnel Output

The same leads and the same spend — with better execution applied.

This section reruns the funnel using your improved rates from Step 3. Nothing else changes — not your marketing spend, not your lead volume, not your offer. The three figures highlighted in green are the commercial result: incremental new members, your improved cost per acquisition, and the percentage reduction in CPA. These are the numbers your leadership team will ask for.

Booked Visits

—

Attended Visits

—

New Members

—

Incremental Members
—
From same spend

New CPA
—
Improved

CPA Reduction
—
Efficiency gain

Step 5 · Revenue & Profit Impact

Incremental members converted into revenue across their full membership lifetime.

Step 4 speaks in members. This step translates members into money. Monthly revenue shows the immediate recurring impact from the additional members gained. Lifetime revenue is usually the number that reframes the conversation — it shows what those incremental members are worth over their full relationship with the business, which is almost always larger than people expect. Gross profit is optional but useful if you want to present the case in terms of bottom-line contribution rather than top-line revenue.

Monthly Revenue Uplift
—

Lifetime Revenue Uplift
—

Gross Profit Uplift
—
If margin entered

⚡ Sensitivity Test — What if uplift is half?

The downside case — automatically calculated at half your projected improvement. No inputs needed. Green = still positive. Red = improvement disappears.

This section cuts your Step 3 assumptions in half and reruns the model automatically. If the results appear in green, the business case holds even under a downside scenario — the decision is structurally sound. If the results appear in red, the business case depends heavily on hitting the full improvement, which should prompt a more conservative initial scope or a phased investment approach before presenting to leadership.

Incremental Members

—

Downside

Lifetime Revenue

—

Downside

CPA

—

Downside

Decision Summary

Turn the numbers into a decision before moving to the Rollout Scorecard.

The business case worksheet produces numbers. This section produces a decision — and those are different things. Teams that complete the financial modelling but leave the strategic logic implicit tend to relitigate the same questions at every subsequent review. Naming the primary constraint, the first lever to move, the biggest risk, and the expansion condition here means that when you reach the Scale Decision Gate in Sheet 5, you are measuring against a pre-agreed standard rather than restarting the conversation.

Most realistic lever

Biggest risk

Expansion condition

Controlled Rollout Scorecard

Complete before go-live. Every field prevents a predictable failure mode.

What this page is for

The business case told you what fixing your funnel is worth. This page decides how you deploy with control.

Most AI deployments that disappoint do so not because the technology failed but because the rollout had no defined boundaries, no named owners, and no agreed standard for success. This scorecard exists to prevent that. Every section corresponds to a failure mode that operators have encountered in practice — scope that drifted quietly, escalation rules that were never written down, ownership that was shared between three people and therefore belonged to none of them.

Complete this before go-live — not after. The decisions you make here are far easier to get right at the planning stage than to correct once a live system is handling real prospect conversations. Work through each section in order. Where a field is unclear, treat that as a signal that the deployment is not yet ready to launch.

The Governance Readiness Checklist at the bottom is your final gate before going live. Every item ticked is a failure mode removed.

2.1 Initial Deployment Scope

Define precisely what is — and is not — included in the first 30 days. The exclusions field is as important as the inclusions.

Sites included ?

Channels included ?

Appointment types ?

Languages / regions ?

Explicit exclusions ?

2.2 Success Thresholds

Set the bar before go-live. A rollout without predefined thresholds becomes a debate. These become the pass/fail criteria at the Scale Decision Gate.

KPI	Baseline (from Sheet 1)	Target

2.3 Escalation & Boundary Rules

A system that escalates when uncertain is a system that stays trustworthy. These rules are how you make autonomy safe.

High-intent signals ?

Always escalate ?

Uncertainty behaviour ?

Max follow-up touches ?

Stop rules ?

2.4 Ownership & Governance Cadence

One name per role. Not a team. Not a role title. A name. Shared accountability is no accountability.

Commercial outcomes owner ?

Performance hygiene owner ?

Knowledge & brand voice owner ?

Integrations & reliability owner ?

Weekly review day & time ?

2.5 Change Control & Reversibility

Who can approve changes ?

Where changes are logged ?

Rollback plan ?

Governance Readiness Checklist

Tick each box only when you can honestly confirm it is in place — not planned, not assumed. If any item is unclear, fix it before expanding.

Scale Readiness Decision Gate

Run after ≥4 weeks live. Expand only when all three sections pass.

What this page is for

This page answers one question: are we ready to scale, or do we need more time?

Every preceding sheet has been building toward this moment. The Execution Baseline showed where your funnel leaks. The Business Case quantified what fixing it is worth. The Rollout Scorecard defined the boundaries and governance for your initial deployment. This gate is where all of that work produces a clear, evidence-based decision — not an opinion, not a feeling, a decision grounded in four or more weeks of live data.

Run this after at least four weeks of live deployment. A single good week is noise. Four consecutive weeks of consistent improvement means the system is performing under real operating conditions — different staff on shift, different days, different lead quality, a public holiday, the week after a campaign ends. All of those variables have cycled through at least once and the performance held. That is what you are looking for before you expand.

Work through the three sections in order. Answer honestly. The decision at the bottom — expand, stabilise, or redesign — follows directly from your answers. Whatever you decide, complete the Next Steps fields before you leave this page.

3.1 Evidence of Performance

Answer based on live deployment data. Use the Notes column to record the specific numbers behind each answer.

Bottleneck (if any No) ?

3.2 Operational Reliability

Conversational performance and operational reliability are not the same thing. Check both.

Top 3 Operational Issues Found

Issue 1 ?

Issue 2 ?

Issue 3 ?

3.3 Governance Confidence

Check whether the governance designed in Sheet 4 is the governance actually functioning. These are different things.

3.4 Decision

Every preceding sheet has been building toward this moment. Select one option and complete the next steps.

✅

Expand scope now

All three sections pass; performance sustained ≥4 weeks

⏸

Stabilise — retest in 2 weeks

One or more No answers; improving but not yet proven

🔧

Redesign before expanding

Systemic reliability or governance gaps; fix root cause first

Scope of next expansion ?

Owner responsible ?

Next review date ?

🎯

Workbook Complete

Your data has been saved. Check your email for a summary of your results and recommended next steps from Keepme.