# 30-Day Behaviour Change Playbook



> A practical guide to changing how your team works in the first month with GoSmarter — parallel vs. cut-over rollout, shop-floor champions, and first-month KPIs to track.
> 
> **URL:** https://www.gosmarter.ai/docs/30-day-behaviour-change-playbook/

**Date:** 2026-05-11



New software doesn't fail at the technical level. It fails because people keep reaching for the spreadsheet. This playbook tells you how to stop that from happening, in the first 30 days, before habits re-form.

{{< callout flag="note" title="Use this alongside the technical guides" >}}
This playbook covers the people side of your rollout. Pair it with the [Implementation Project Plan](../implementation-project-plan/) for the full technical and project management picture.
{{< callout-end >}}

## The Core Challenge: Software vs. Habit

Every production manager in metals has a personal spreadsheet. Some are masterpieces with years of accumulated logic, colour-coded tabs, formulas that nobody else understands. They work. That's the problem.

GoSmarter needs to become the trusted source of truth before people stop cross-checking it against the spreadsheet. The first month is the window. Get it right and adoption sticks. Get it wrong and the spreadsheet wins.

The playbook is split into three parts:

1. **Rollout approach**: parallel or cut-over?
2. **Champion network**: who leads the change on the floor?
3. **First-month KPIs**: how do you know it's working?

## Part 1: Parallel vs. Cut-Over Rollout

### What the Terms Mean

**Parallel running** means operating GoSmarter and your existing process at the same time for a defined period. Every action that goes into GoSmarter also goes into the old system (or spreadsheet). You compare outputs. When you trust the new system, you stop the old one.

**Cut-over** means picking a date and switching. From that morning, GoSmarter is the only system. The old spreadsheet is archived. You don't look back.

Neither approach is always right. The decision depends on your risk tolerance, your team size, and how different GoSmarter's output is from your current process.

### When to Run Parallel

Choose parallel running if:

- Your existing process has legal or audit obligations (e.g., mill certificate (also called a [Material Test Report (MTR)](/hubs/metals-manufacturing-glossary/#mill-test-certificate-mtc)) filing must remain unbroken during transition)
- You have a large team and can't retrain everyone simultaneously
- GoSmarter is replacing a bespoke internal tool with unusual logic that needs validation
- Your production volume is high enough that a mistake has immediate commercial consequences

**Parallel period recommendation:** two to four weeks. Longer than four weeks and people stop taking GoSmarter seriously as it becomes the "other" system rather than the real one.

**How to run it well:**

- Designate a single person to compare GoSmarter's output with the legacy output each day. This should take no more than 15 minutes.
- Log every discrepancy in a shared document. Categorise as: (a) GoSmarter is wrong, (b) legacy system is wrong, (c) the inputs were different, (d) both are right but show the same thing differently.
- Set a go/no-go date at the start. Don't let it drift. If you reach the date and have unresolved category (a) discrepancies, escalate to GoSmarter support before extending the parallel period.

{{< callout flag="tip" title="Run parallel for mill certificates first" >}}
Even if you cut over on inventory and orders, consider a short parallel period for mill certificate uploads. Regulators and customers don't care about your change management programme — they need the cert. Validate the extraction accuracy on real documents before you rely on it for compliance.
{{< callout-end >}}

### When to Cut Over

Choose a clean cut-over if:

- You have a small team (fewer than 10 users) who can be trained in one session
- Your existing process is genuinely broken, parallel running would mean doubling up on broken work
- You're starting fresh with GoSmarter before a new contract or site opens, so there's no legacy data to compare against
- Your team has already piloted GoSmarter in a limited context and trusts it

**How to run it well:**

- Archive, don't delete, the legacy spreadsheet or system. People need to know it still exists for reference. The goal is to reduce the psychological cost of switching: if they know they can look back, they're less likely to keep both running in parallel anyway.
- Pick a Monday. Never cut over mid-week. You want a clean weekly boundary.
- Do a full data import check the Friday before. If anything looks wrong, you have the weekend to fix it without production pressure.
- Brief every user on the Friday. Not an email, have a conversation, even a five-minute stand-up at shift start.

### The Hybrid Approach

Most medium-sized operations do this in practice: cut over on inventory and orders (low risk, easy to validate), run parallel on production planning for two weeks (higher stakes, needs confidence), and run parallel on mill certificates for four weeks (compliance-critical).

This is sensible. Document your approach explicitly so the whole team knows what's live and what's still being validated.

## Part 2: Nominating Shop-Floor Champions

### Why Champions Matter

You can brief a team once. A champion reinforces the message every day, answers the five-minute questions that nobody sends an email about, and spots when someone has quietly gone back to their spreadsheet.

The best champions aren't always the most senior people. They're the ones who other people ask when something doesn't look right.

### Who to Nominate

Look for people who have:

- **Credibility on the floor**: their colleagues trust their judgement, not just their job title
- **Natural curiosity**: they've already asked questions about GoSmarter or explored it without being asked
- **A production task that GoSmarter directly improves**: if the champion's own day gets better, their advocacy is genuine

Aim for one champion per shift if you run multiple shifts. If you have separate teams for inventory, production planning, and mill certificates, ideally nominate a champion in each area.

**Avoid nominating:** the most senior person in the room by default, anyone who openly resisted the rollout decision, or anyone with a workload that leaves no time to support colleagues.

### What Champions Do

The champion role is not a second job. Define it narrowly:

| Task | Time commitment |
|---|---|
| Attend a 30-minute champion briefing before go-live | Once |
| Be the first point of contact for questions from their team | As needed |
| Do a five-minute daily check of GoSmarter during week one | Week 1 only |
| Flag recurring issues or confusion to the implementation lead | Ongoing |
| Join a brief weekly check-in call with the implementation lead | Weeks 1–4 |

{{< callout flag="note" title="Champion briefings" >}}
GoSmarter's customer success team can join your champion briefing to run a Q&A session. Contact us via the [GoSmarter app](https://app.gosmarter.ai) or book a session through [your account dashboard](https://app.gosmarter.ai).
{{< callout-end >}}

### Supporting Your Champions

Champions need three things to succeed:

1. **Authority**: a clear message from the implementation lead (or site manager) that the champion's guidance should be followed. Without this, champions get ignored.
2. **An escalation path**: when a champion can't answer a question, they need to know exactly who to ask and how quickly they'll get a response. Document this on day one.
3. **Recognition**: at the end of the first month, acknowledge what the champion did. It doesn't need to be formal. A mention in a team meeting is enough. People remember these things.

## Part 3: First-Month KPIs

You need to know whether the rollout is actually working, not just whether people are logging in. These three KPIs tell you what you need to know in the first 30 days.

### KPI 1: Shadow Spreadsheet Usage

**What it measures:** Whether your team is still maintaining parallel records outside GoSmarter.

**How to track it:** Ask each champion to do a 10-minute informal check at the end of weeks two and four. Are people keeping personal spreadsheets updated? Are separate tracking sheets still circulating via email or WhatsApp?

This is a behavioural signal, not a system metric. You won't get it from a dashboard. You get it from asking.

**What good looks like:** By the end of week four, no active spreadsheets covering data that GoSmarter now holds. Legacy sheets exist for reference only and haven't been updated in at least two weeks.

**What to do if it's not good:** Don't shame the spreadsheet users. Find out what GoSmarter isn't doing that the spreadsheet was doing for them. Is it a missing field? An export they can't get? A report format they need? Fix the root cause, not the behaviour.

### KPI 2: AI-Plan Acceptance Rate

**What it measures:** Whether planners are trusting GoSmarter's optimised cutting plans or overriding them every time.

**How to track it:** GoSmarter logs plan acceptance and overrides in the optimisation output. Review this in your Dashboard at the end of weeks two and four. Look at the ratio of accepted plans to manual overrides.

**What good looks like:** An acceptance rate above 70% by the end of week four. 100% is not the goal. Experienced planners will correctly override the AI when they have information GoSmarter doesn't (a machine is down, a customer has a specific preference). You want the AI to be the starting point, not something people discard without reading.

**What to do if it's not good:** Low acceptance usually means one of three things:

- The input data (stock levels, order details) is inaccurate, so the plan is generating wrong results
- Planners don't understand what the plan is optimising for and don't trust the logic
- The plan format is unfamiliar and the planner is defaulting to what they know

Talk to your champion and sit with a planner for 30 minutes. You'll identify the issue quickly.

{{< callout flag="tip" title="Track this weekly, not just at the end of the month" >}}
If acceptance is low in week one, that's normal. The AI is learning your data and your planners are learning the format. If it's still low in week three with no upward trend, escalate early.
{{< callout-end >}}

### KPI 3: Certificate Search Time

**What it measures:** How long it takes a user to locate a specific mill certificate (MTR) in GoSmarter vs. how long it used to take.

**How to track it:** Ask three or four users who regularly retrieve certificates to time themselves on a real search in week one, then again in week four. You want before-and-after data from the same people on the same type of task.

A baseline search in a paper or folder-based system typically takes five to twenty minutes. GoSmarter search should return results in under 30 seconds once certificates are uploaded.

**What good looks like:** Search time under two minutes by week four, including any time spent interpreting the result. The two-minute target accounts for users who are still building familiarity with the interface.

**What to do if it's not good:** The most common reason search is slow is that certificates haven't been consistently uploaded or tagged. Check your upload completeness rate (what percentage of recent deliveries have a certificate in GoSmarter?) and address any gaps before assuming the search tool is the problem.

## Week-by-Week Timeline

| Week | Focus | Champion actions | Implementation lead actions |
|---|---|---|---|
| Week 0 (pre-go-live) | Champion briefing, data import check, archive legacy systems | Attend briefing, read this playbook | Confirm data imports, pick KPI baselines, brief site manager |
| Week 1 | Go-live, daily check-ins | Five-minute daily check, answer questions, log issues | Daily stand-up with champions, monitor plan acceptance |
| Week 2 | First KPI check, address blockers | Shadow spreadsheet check, flag recurring issues | Review acceptance rate, check search times, fix root causes |
| Week 3 | Normalisation | Champion available for questions, not actively monitoring | Confirm parallel period end date (if running parallel) |
| Week 4 | End-of-month review | Final shadow spreadsheet check | Review all three KPIs, document what worked, share with GoSmarter |

## Escalation: What to Do When It's Not Working

If you reach the end of week four and adoption is poor, run through this checklist before concluding that the software isn't right for your team:

- **Is the data accurate?** Poor data produces poor outputs. If GoSmarter's inventory doesn't reflect actual stock, nobody will trust its plans.
- **Did every user get a hands-on training session, not just a walkthrough?** Watching a demo and doing the task are different. Require that every user completes at least one real task in GoSmarter before go-live.
- **Has the implementation lead checked in with non-champions?** The champions give you signal, but they're also optimistic by nature. Speak directly to the most resistant users.
- **Is there a specific feature gap?** Log it with GoSmarter support. Many feature requests come from early adopters in the first month.

{{< callout flag="note" title="GoSmarter support" >}}
If you're struggling with adoption, contact us. We know change is hard and have worked with dozens of companies. Reach out to us via [talktous@gosmarter.ai](mailto:talktous@gosmarter.ai) or reach out to the person currently supporting you.
{{< callout-end >}}

## Summary: Your First-Month Checklist

- [ ] Decided on parallel vs. cut-over approach and documented it for the team
- [ ] Named at least one champion, confirmed their authority and escalation path
- [ ] Set a baseline for all three KPIs before go-live
- [ ] Archived (not deleted) all legacy spreadsheets and tools
- [ ] Completed week-two KPI check and addressed any blockers
- [ ] Completed week-four KPI check and shared results with GoSmarter