Change and adoption
·
7 min read
Key findings
A pilot runs on volunteers, attention, a watching sponsor and exemptions, and a rollout has none of them.
Pilot purgatory is the state in which nobody decides to scale a pilot or stop it.
The scale criteria are agreed with the business owner before the pilot: measure, baseline, threshold, second site, date, budget and owner.
The second site should be ordinary, with a manager who did not volunteer and a normal workload.
An HR team can run pilots indefinitely at no cost to the business, so the business is never forced to choose.
Pilots fail to scale because the conditions that produced the result do not travel.
A pilot runs on volunteer managers, extra attention, a sponsor who is watching and exemptions from competing work, and a full rollout has none of the four.
Scaling depends on agreeing with the business owner, before the pilot starts, what the second site must show.
Ashkenas and Matta open their Harvard Business Review article with the observation that successful pilots often do not lead to wider implementation.
Four Conditions A Pilot Enjoys
A pilot succeeds partly on merit and partly on conditions that nobody lists at the time.
Volunteers. The managers who took part wanted the practice and behaved accordingly, whereas a rollout includes managers who never asked for it
Attention. The practitioner ran the sessions, answered questions the same day and fixed problems as they appeared
A watching sponsor. An executive asked about progress, which changed how local managers spent their time
Exemptions. Pilot units were excused from a report, a target or a competing project
A fifth condition often sits beside these, which is a ring-fenced budget.
Scope is a sixth.
Sparkman notes that a workforce planning pilot involves fewer dependent stakeholders, less data and a smaller scope than the enterprise version that follows.
None of these conditions is available to a unit that receives the practice through a rollout announcement.
A pilot that relied on them shows that the practice works under conditions the organisation cannot reproduce.
Pilot Purgatory
Pilot purgatory is the state in which a pilot is extended, repeated in a new unit or relaunched, and nobody decides to scale it or stop it.
Four signs show it early.
The pilot has run for longer than its original plan
A second pilot has started in a unit that resembles the first
The sponsor asks for more evidence and does not say what evidence would settle the question
Nobody outside HR has a budget line for the rollout
The cause is usually the absence of a scale decision, and the missing decision belongs to a business owner.
An HR team can run pilots indefinitely at no cost to the business, so the business is never forced to choose.
Scale Criteria Before The Pilot
The most useful work happens before the pilot begins, in a conversation with the executive who would pay for the rollout.
Six things are agreed in that conversation and written on one sheet.
The measure that shows the practice worked, taken from data the business already holds
The baseline for that measure, and the improvement that would justify scaling
The named ordinary second site, chosen before the results are known
The date of the scaling decision, and the person who takes it
The budget the owner would release if the threshold is met
The owner of the rollout once the HR project team steps back
Agreeing the threshold in advance stops a marginal result being reread as a success once the momentum exists.
An owner who has signed the sheet has also accepted that a missed threshold means the practice stops or changes.
Evidence A Business Owner Funds
A business owner deciding whether to fund scale looks at four things.
The effect in the ordinary second site, against the baseline
The full cost per unit at scale, including the manager time the practice needs each month
The capability the practice requires, and whether the organisation has it without the practitioner
The name of the person who owns it after the pilot team leaves
A satisfaction score from pilot participants answers none of the four.
Report the four as one page with the threshold beside the result.
Choosing The Second Site
The instinct is to choose a second site likely to succeed, which produces two successes and no new information.
The useful second site is ordinary.
Its manager did not volunteer, is competent and unenthusiastic, and carries a normal workload.
What happens there is what will happen across the organisation, and it happens while the programme is still small enough to change.
Three outcomes are all useful.
The practice works, which is evidence and not a demonstration
It works with changes, which gives the version that scales
It fails, which is the cheapest failure available
Stripping The Pilot's Support
A pilot accumulates support that cannot be replicated, and scaling means removing it on purpose.
Four things usually have to go.
The practitioner's personal involvement, replaced by a written standard
The bespoke materials, replaced by a version a manager can use without help
The extra measurement, reduced to what the business already collects
The exemptions, since the practice has to sit beside normal work
Stripping these usually reduces the effect.
Running the second site without them shows by how much before the rollout is announced.
A satisfaction score from pilot participants answers none of the four.
A Worked Example
Consider a hypothetical manufacturer that pilots a structured onboarding programme in two plants.
Both plant managers volunteered, the HR business partner ran the first-week sessions herself, and the operations director asked for monthly updates.
Ninety-day retention of new starters rose from 82 per cent to 91 per cent in the two pilot plants.
Before the pilot, HR and the operations director had agreed the measure and a threshold of 86 per cent.
They had also named a third plant as the second site, chosen because its manager was competent and had shown no interest.
At the third plant HR provided a written checklist and no facilitator.
Retention there rose from 82 per cent to 87 per cent, which cleared the threshold by one point.
The operations director released a budget line and named the regional head as owner of the rollout.
HR handed over the checklist and a ninety-day review date, and stepped back.
All the figures are illustrative.
The decision was made against a threshold that existed before the result did.
Mandated, Supported Or Available
Rollout is often announced as a decision and carried out as an invitation.
Three positions exist, and confusing them is a common failure.
Mandated. The practice is required, reported and audited, which suits anything with a compliance or safety consequence
Supported. Units choose, and the centre provides materials and help, which suits practices whose value is local
Available. The materials exist and nobody follows up, which makes the practice a library item and not a programme
Say which one applies before the announcement.
The three need different resourcing and produce different reporting.
Adapt Or Replicate
A rollout does not have to copy the pilot exactly.
The Transportation Security Administration built its conflict management system by starting where the energy was.
Eleven airports that had volunteered as pilot sites joined a design workshop with cross-functional teams.
Their proposal fixed a core, which was conflict management training for all staff and a communication strategy.
It then offered local options, such as peer review panels, employee councils and confidential concern forms.
The external consultant urged flexibility, and the core concepts could be applied differently at each site.
At the Milwaukee airport, supervisors report each month on the principles they have put in place, and the reports are shared with all supervisors.
Later, conflict competencies were reviewed in relation to hiring, promotion, performance management and rewards.
The case shows a volunteer start followed by a deliberate move into the rules that govern ordinary work.
What Scales And What Does Not
Practices that scale are usually a changed routine, a written standard or a decision moved to a different level.
Practices that do not scale usually depend on an individual's skill, a relationship or sustained attention from the centre.
Where the pilot's effect depended on a skilled facilitator, scaling needs either many skilled facilitators or a redesign that removes the need.
Naming which of the two is proposed makes the resourcing conversation honest.
Who Owns The Rollout
The rollout needs an owner in the line, named on the criteria sheet and not appointed after the decision.
The owner writes the rollout plan, holds the budget and reports adoption to their own manager.
HR supplies the standard, the materials and the measurement, and its role in the plan is written down.
A rollout owned by the HR project team ends when the project team is reassigned.
The Middle Layer
Rollouts stall where programmes usually stall, in the layer of managers who deliver the practice.
They are asked to deliver something they did not design, without control over how, while their own results are judged this quarter.
The four conditions behind that stall are set out in why change programmes stall at the middle.
A rollout that changes a manager's target or removes a task before it adds one meets far less silence.
Measuring Adoption
Rollout reporting counts units covered, which says nothing about whether anything changed.
Three measures say more.
Use without prompting, after the launch support ends
Local adaptation, which indicates ownership
The original measure, in the new units
A unit that has changed the practice to fit its work has adopted it.
A unit running it exactly as issued may be complying.
Adoption also falls after launch support ends.
The practice survives better when it is attached to something that already recurs, such as an existing meeting or form.
One check at ninety days, with a named owner, is enough.
Stopping A Rollout
Some pilots should not scale, and saying so is the most valuable use of the evidence.
Three findings justify stopping.
The effect disappeared in the ordinary second site
The practice requires a capability the organisation does not have and will not build
The cost at scale exceeds the cost of the problem
A function that has stopped one rollout on evidence is believed about the next one.
Related Reading
A programme that changes the rules behind behaviour, as in changing culture by changing rules, needs the same scale criteria.
Where the pilot tested a new HR measure, people analytics without a data team covers the evidence standard.
For the business case behind the funding request, see the business case for a people programme.
Where The Tools Sit
Trusted Talent Advisor covers the adoption sequence and the conditions work in its change and adoption component.
The measure and threshold sit in its evidence and impact component, and the template pack includes the adoption tracker and the pilot brief.
Sources
1. Ashkenas, R. and Matta, N. How to Scale a Successful Pilot Project. Harvard Business Review.
2. Runde, C. E. and Flanagan, T. A. Developing Your Conflict Competence: A Hands-On Guide for Leaders, Managers, Facilitators, and Teams. Jossey-Bass.
3. Sparkman, R. Strategic Workforce Planning: Developing Optimized Talent Strategies for Future Growth, second edition. Kogan Page.
From the system
Trusted Talent Advisor sets out the full method: seven components, 51 tools, The Talent Template Pack and The Prompt Library.
See the system
Questions this answers
Why do successful pilots fail to scale?
The pilot ran on volunteer managers, extra attention, a watching sponsor and exemptions from competing work. A rollout has none of these, and nobody agreed beforehand what would justify funding it.
What is pilot purgatory?
It is the state in which a pilot is extended, repeated in a new unit or relaunched, and nobody decides to scale it or stop it. It persists because the business owner never agreed a threshold for the decision.
How do you know a pilot is ready to scale?
The practice met the threshold agreed beforehand, in an ordinary second site, without the practitioner's involvement or exemptions. A named owner in the line holds a budget and a plan for the rollout.
Should the rollout copy the pilot exactly or let teams adapt it?
Fix a core and allow local options. A unit that adapts the practice to its work has adopted it, and a unit running it exactly as issued may be complying. The core is what the measure checks.
How do you design a pilot so the business will fund the rollout?
Agree the measure, baseline, threshold, ordinary second site, decision date, budget and rollout owner with the business executive before the pilot begins, and write them on one sheet.
Who owns a rollout once HR steps back?
A named manager in the line, agreed on the criteria sheet before the decision. The owner writes the plan, holds the budget and reports adoption to their own manager, and HR supplies the standard and the measurement.
Keep reading
Measuring HR impact
Internal Mobility Metrics Without A Platform
Seven measures an HR adviser can build from existing records, each with a stated definition, a base and a way to read it.
7 min read
Workforce and future skills
Job Rotation As A Talent Gap Strategy
A job rotation programme closes a talent gap when the role, the person, the cover and the evidence are agreed with both managers before it starts.
8 min read
Talent reviews and succession
Seeing Talent Below The Top Team
Visibility of leadership potential drops below the top team. A cross-unit forum, run on one page per unit and a written evidence standard, is the method for seeing it.
8 min read