Why New HR Initiatives Don't Stick

"We did something like this two years ago. It ran for about four months."

Somebody senior said a version of that when you announced the new programme. Not as an objection — an objection would have been easier. It was said the way people report the weather.

The programme had a name. It had a launch, a deck, a WhatsApp group, a champion in each function and a tracker HR updated every Friday. By month two, attendance on the review call had thinned. By month four, the tracker was being filled in from memory. By month six, nobody mentioned it except HR, and HR mentioned it because they were being asked to.

You have been through this before. A values refresh. A new appraisal format. An engagement survey with an action plan. Each consumed real money and, more expensively, a quarter of your senior team's attention.

Two claims here. The first is about the statistic you were quoted when you bought the programme, which has no evidence behind it. The second is about why yours actually failed, which has nothing to do with communication, buy-in or change management.

Consequence is the fourth of the five conditions in employees not taking ownership, and it is the condition an initiative is meant to supply. Most initiatives never supply it.

The number every consultant opens with

Somewhere in the first ten minutes of the pitch, you were told that 70% of change initiatives fail. It is the most quoted figure in corporate change work, and it appears on almost every page that ranks for this question.

There is no established evidential basis for it.

Mark Hughes of the University of Brighton went looking for that basis and published what he found in the Journal of Change Management in 2011, in a paper titled "Do 70 Per Cent of All Organizational Change Initiatives Really Fail?". He took five published instances of the claim and traced each back to whatever it rested on. In every one, the supporting evidence was absent.

The traceable origins of the figure are these.

• **Hammer and Champy, Reengineering the Corporation, 1993.** They offered an unscientific estimate of "50 to 70 percent". Hammer said later that it had been misrepresented — his point had been descriptive, not a failure rate.
• **Beer and Nohria, Harvard Business Review, 2000.** They stated 70% flatly, with no reference at all.
• Kotter, 2008. An estimate drawn from personal observation.
• McKinsey, 2009. A survey of 1,546 executives in which 30% rated their programmes completely or mostly successful. This is the closest thing the figure has to a source, and it is a survey of executive opinion — not a measured failure rate. Executives were asked how they felt their programmes had gone, and they answered.

Be careful about what this does and does not establish, because the overclaim is tempting and wrong. It does not establish that change programmes usually succeed. It does not establish any number at all. The finding is narrower and firmer: there is no reliable empirical basis for the specific figure being quoted to you as though there were one.

Which matters commercially, in a particular way. A number with no evidence behind it is being used to sell you something. The same number is then available, afterwards, to explain why the thing you bought did not work. The failure was priced into the pitch, and a vendor whose category admits to a 70% failure rate before the engagement begins is never the party at fault when the engagement fails.

So put the number down. It tells you nothing about your business. What happened inside your business is knowable.

The people who own the system do not believe in it either

Here is a figure that is properly sourced.

Gallup surveyed 135 Fortune 500 CHROs in 2023 and found that 2% strongly agree their performance management system inspires employees to improve. Two percent. Not employees complaining about appraisals — the chief HR officers of the largest companies in the world, describing the system they chose, paid for and are accountable for.

The employee side agrees with them, which is what makes the pair credible. Gallup's 2023 research found 22% of employees strongly agree their performance review process is fair and transparent, and roughly one in five say reviews inspire better performance.

The most widely installed people system on earth is not believed in by the people who own it, is not believed in by the people it is applied to, and continues to be installed everywhere including here. Your initiative did not fail because you are behind. It failed for a reason that is structural and visible from your chair.

What actually killed yours

The standard explanations are poor communication, insufficient buy-in, weak change management and lack of leadership alignment. You will have heard all four, probably in the review that followed the programme going quiet.

Every one of them is unfalsifiable. There is no quantity of communication that would let you declare communication sufficient, and no measurement of buy-in that could disprove the buy-in explanation. Each is offered after the fact, cannot be tested, and — notice this — implies that the remedy is more of what you already bought.

Your programme failed a different test, within about three weeks of launch. Everybody it applied to ran it, quietly, without discussing it. One question.

After the programme launched, did anything different happen to any individual as a result of behaving in the new way, or as a result of behaving in the old way?

If nothing did, the programme was an announcement. People correctly identified it as one and waited it out. They are good at waiting things out, because the last three initiatives also went away and waiting cost them nothing on any of those occasions. That is not cynicism. It is an accurate model of your organisation, built from evidence, and it has been right every time so far.

The Consequence Test

Four questions. Ask them before you launch anything, and of anything currently running that you are unsure about. They take an afternoon and they are unpleasant in the right way.

Who is measurably better off if they adopt it — and when do they find out?

Not who benefits in principle. Who, by name, ends up better off in a way they can point at, and on what date.

Vague future benefit is not a consequence. "This will make the company stronger" is not one. A consequence is money, scope, a decision right, a title, a statement made in a room that matters — something a specific person receives at a time they can predict.

Timing is where most initiatives die unnoticed. Gallup's 2023 data found that 56% of employees formally review goals with their manager once a year or less. If the only moment your organisation attaches an outcome to behaviour is the annual appraisal, a programme launched in July has its first real consequence the following March, assuming it is still alive. Nothing survives an eight-month gap between behaviour and result.

Who is measurably worse off if they ignore it — and has that actually happened to anyone by name yet?

Until it has, the answer is nobody. Everyone knows this, including the people telling you the rollout is going well.

Organisations do not read policies. They read cases. The first person who visibly ignored the new process and had nothing happen to them is now the policy, and the policy is that this is optional. That person is usually senior, often the highest biller, sometimes family. The exemption feels small and pragmatic when you grant it. It is the entire programme, decided in one conversation you have probably forgotten.

Consequence that only ever runs downward produces caution rather than adoption — the same asymmetry we take apart in the accountability gap.

What did it replace?

An initiative that adds to the existing workload without removing anything is a tax. Taxes get avoided, quietly and competently, by capable people with other work to finish.

The new form went on top of the ERP entry. The competency framework went on top of the KRA sheet nobody had updated since the company was doing ₹18 crore. Nobody's week got shorter. The people most likely to comply are the ones with the least real work, which is exactly the wrong selection.

Name what stops. If nothing stops, you have not launched an initiative — you have raised the cost of producing the same output, and your best people absorb that cost first and longest. What that does to a good performer over two quarters is in why good employees underperform.

Did the promoter's own behaviour change?

This is the decisive one, and it is worth more than the other three combined.

Everyone in an Indian promoter-led business calibrates on the promoter. Not on the policy, not on HR, not on the professional CEO if there is one. On what the promoter asks for, what he escalates on, and what he praises. Those three signals are read far more carefully than any document.

So if the initiative says decisions should be made closer to the work, and in Monday's review you ask the plant head why he approved a ₹2 lakh spend without checking with you first, that question has outranked the entire programme. It took nine seconds, and nobody in the room needed to hear the rest of the meeting. The founder's half of this pattern is in why founders micromanage; the structural version — writing decision rights down so authority stops depending on mood — is in the decision bottleneck.

How to test it: take the last six weeks of your own WhatsApp messages and emails to your top team. Count how many relate to the initiative you launched. Then count how many relate to what you were chasing before it launched — collections, that one client, yesterday's dispatch number. You will not need a spreadsheet to see the ratio. It takes forty minutes, alone.

Why nobody asks the fourth question

Because the initiative was delegated to HR.

HR can run the launch, build the tracker, chase the champions, design the form and write the reminders. HR cannot change what the promoter asks for in Monday's review. So the function held responsible for the outcome has authority over everything except the single variable that determines it — and carries the failure, usually in the same meeting where the programme is declared to have lacked buy-in.

It is not that your HR head does not know this. Ask one privately and you will get a precise answer in under a minute. Very few have the standing to say to a promoter: this will not work unless you change what you ask for on Monday. The ones who say it once and are not heard do not say it twice.

That is responsibility without authority, the same failure that hollows out a middle layer — see why middle management fails. So the question that decides the outcome belongs to the one person in the building nobody will put it to. Which means you have to put it to yourself.

The alternative, which will look disappointingly small

One change. One team. One quarter.

Before you announce anything, write four things down. What the person who adopts it gets, and the date. What happens to someone who does not, and who delivers it. What stops, to make room. And what you will stop asking for, start asking for, and start praising — named specifically, in the same document, as part of the change.

That last line is not a courtesy. It is the change. If you cannot write it, the other three will not hold — better to know that before the launch than in month four.

No name. No logo. No launch event. You announce it in one meeting, to one team, in about four minutes, and then you behave differently for ninety days where people can see you.

This is the pillar's argument applied to programmes: ownership fails at the lowest missing rung, and repairing that one rung in one team beats launching anything across the company. Smaller, far less impressive to describe to a board, and the only version that survives a normal working week — because it is the only version with a consequence attached.

Where to start

Do not launch anything this quarter.

Take the initiative you launched most recently, the one still nominally alive, and run the four questions against it, on paper, alone, in an afternoon. Most promoters stall at question two, discover the exempted person, and recognise them immediately.

If you reach question four and the honest answer is that nothing about your own week changed, you know why the programme died — and that no relaunch, better communication plan or new vendor addresses it.

If you want to see the pattern across the whole business first, the Business Pulse diagnostic maps where decisions and consequences actually sit in your organisation, as against where the process document says they do.