Software goes live on a date. Work changes over the following months. The method you choose determines whether that change is planned, measured and reinforced, or left to the loudest manager and the fastest training session.
This comparison looks at five approaches a Bangladeshi SME can realistically run during a software implementation. They are not mutually exclusive, but they lead to different plans, different budgets and different evidence of adoption.
Five methods, one adoption problem
- Sponsor-led programme. An executive sponsor and a dedicated change lead run a structured plan across functions, with a governance group and a benefits review.
- ADKAR. A readiness model that works through awareness, desire, knowledge, ability and reinforcement, diagnosing where each affected group is stuck.
- Kotter's eight steps. A campaign sequence that builds urgency, a coalition, a vision, communication, empowerment, short wins and consolidation.
- Agile embedded adoption. A product owner and user champions iterate the rollout, release in slices and adjust the process from feedback.
- Training-led rollout. A curriculum, super-users and go-live floor support, with adoption assumed to follow competence.
Eight criteria for comparing adoption methods
Compare the methods on the same dimensions, or the choice becomes a matter of preference.
- Sponsorship depth: how explicitly the method forces executives to act, not just approve a budget.
- Readiness diagnosis: whether it helps you find which groups lack awareness, desire, ability or reinforcement.
- Resistance handling: how it treats legitimate disagreement, process exceptions and informal power.
- Manager role: whether line managers are given scripts, data and accountability.
- Adoption measurement: whether it produces behavioural evidence rather than attendance sheets.
- Speed to first use: how quickly real users complete real work in the new system.
- Effort and skills: what it demands from a team that already has a day job.
- Durability: whether reinforcement continues after the launch event.
How the five methods compare
| Approach | Strongest at | Weakest at | Best fit | Effort |
|---|---|---|---|---|
| Sponsor-led programme | Cross-functional decisions, escalation and visible commitment | Can become governance theatre without floor-level reinforcement | High-risk, multi-department replacements | High |
| ADKAR | Diagnosing where a specific group is stuck | Needs disciplined measurement; weak as a standalone communications plan | Mixed readiness and behaviour gaps | Medium |
| Kotter eight steps | Building urgency and a coalition for a long programme | Easily reduced to a launch campaign; slower to start | Low urgency, large organisational change | Medium to high |
| Agile embedded adoption | Fast feedback, process fit and iterative release | Can under-invest in formal readiness and manager accountability | Product teams and phased rollouts | Medium |
| Training-led rollout | Competence and confidence on a narrow tool | Assumes willingness and a stable process; weak on resistance | Small, low-risk tool changes | Low to medium |
What each method looks like in a mid-sized firm
Sponsor-led means the managing director chairs a fortnightly adoption review, department heads bring completion data, and blocked decisions are resolved in the room. The change lead maintains a plan, a risk log and a resistance map. The risk is that the programme becomes a reporting exercise; the test is whether a supervisor changes a rota, an incentive or a workaround because of it.
ADKAR starts with a short diagnostic per group. If accounts staff understand the new approval flow but keep using the old spreadsheet, the gap is ability or reinforcement, not awareness. The plan then targets the specific barrier: practice time, a simplified screen, a manager check or a removed legacy shortcut.
Kotter is useful when the organisation does not yet believe change is necessary. It invests in a coalition, a clear story and visible short wins. In practice, this can look like a weekly floor walk, a named champion in each team and a public before-and-after metric. Without a plan to anchor the change, momentum fades after the launch lunch.
Agile embedded adoption releases the system in slices, watches how people actually use it and adjusts the workflow. It suits implementations where requirements are still being discovered. It needs a product owner with authority and a feedback loop that reaches the vendor quickly, otherwise iteration becomes churn.
Training-led is the most common default. It works for a narrow change with a stable process and a willing user group. It fails when the real barrier is a changed incentive, an unresolved data-quality problem or a manager who quietly keeps the old process alive.
Choose the method by risk and readiness
- Sponsor-led when the change crosses departments, affects money or compliance, or has failed before.
- ADKAR when you can name the behaviour gap but not the reason for it, especially across several user groups.
- Kotter when urgency is low and the programme will take long enough that coalition-building pays back.
- Agile embedded when the process is still evolving and a phased release is technically feasible.
- Training-led when the tool is narrow, the process is stable and resistance is genuinely low.
- Hybrid for most SMEs: sponsor-led governance, an ADKAR diagnostic, training for ability and line-manager reinforcement after go-live.
Score the choice before you commit
Give each criterion a weight from one to three based on the risk of getting it wrong. Score each method from zero to three against evidence, then multiply and add. Keep the evidence beside the score: a named sponsor, a readiness survey, a manager commitment, a pilot result. A weighted score does not make the decision for you, but it exposes which assumptions are carrying the conclusion.
Mistakes that make any method fail
- Treating communication as change management.
- Training people before they understand why the change matters.
- Leaving line managers out of the reinforcement plan.
- Measuring attendance instead of completed work in the new system.
- Giving the change team responsibility without authority to unblock decisions.
- Ignoring the workarounds that keep the old process alive.
A one-page change plan
Name the sponsor, the change lead and the managers accountable for adoption. List the behaviour changes, the groups affected, the barriers, the interventions, the dates and the measures. Add a review cadence and the decision that will be taken if adoption misses the target. If the plan needs more than a page, it is probably a programme description rather than a plan.
If you need help sequencing the rollout or building adoption measures into the implementation, start with Custom Software & Integrations. Our pricing page covers how discovery, build and post-launch support are scoped. For the wider people side, read how AI adoption changes year by year in an SME.
Sources
- Prosci, The ADKAR Model, accessed 17 November 2026.
- Kotter, The 8 Steps for Leading Change, accessed 17 November 2026.
- Project Management Institute, Change management learning resources, accessed 17 November 2026.
- ISO 9001: Quality management systems — Requirements, accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.