Push and Absorb: A Framework for Making Change and Meeting It

The idea in one sentence

Whatever you are forced to deal with was, at some point, something somebody else chose to do.

A fuel price rise is a decision taken in a ministry. A rival’s discount is a decision taken in a boardroom. A new rule is a decision taken by a regulator. None of it fell out of the sky. It reached you as a condition, and it left somebody else’s desk as a choice.

This runs both ways. The change you make this quarter lands on a supplier, a worker or a customer as a condition they did not pick.

So handling change is two skills, and they mirror each other.

The vocabulary is the same in both columns and the direction of travel is opposite. Run them as one exercise and you get a checklist that fits every situation and settles none of them.

The practical aim is to spend more of your working life in the left column. Every firm does some of both. The strong ones have shifted the ratio.

The first question: do I hold the lever?

Before any analysis, ask whether this change is yours to make or yours to take.

• Yes, I hold it. Run Push.
• No, somebody else holds it. Run Absorb.
• Partly. Split the situation in two and run one framework on each half.

A lever here means a decision you can take and act on this week. Price, staffing, incentives, what you sell, where you sell it. Things you only influence, like what people think of you, do not count.

When the answer is “partly”, split the situation before you go any further. Milk prices rising is not mine. What I charge for tea is mine. Two problems, two methods.

Mixing them fails in two directions. You treat a decision you could simply take as something to be managed, and lose a quarter to meetings. Or you treat a market force as something to be fought, and spend money on a battle you were never going to win.

Answer the lever question honestly rather than hopefully. Plenty of things people call uncontrollable are only uncomfortable, and plenty of things people claim to control belong to somebody senior to them.

Framework A. Push: change you start

Four moves, in order: define, trace, commit, measure. Then start again.

1. Define the change

A real change answers four questions. If you cannot answer all four, what you have is an intention.

Write it as one sentence. Move X from A to B, for these people, paid for by these people, by this date.

The rule that does the most work in this document:

Good and bad are not properties of a change. They are properties of three things together: the change, the person you are asking about, and how far ahead you look.

Cutting a staff bonus is good for this month’s margin, bad for next quarter’s staffing, and bad for the staff on every horizon. All three are true at once. A change becomes “good” only once you have said whose seat you are sitting in and what you are trading away.

Arguments about whether a change is good usually skip both of those, which is why they go nowhere.

2. Trace what it touches

First, how does the change travel? Your decision rarely touches the result directly, and the path is where changes die.

Paths that run through belief fail quietly. Nobody tells you it did not work. The number simply never moves.

Second, what gets knocked over? “Think about knock-on effects” gives you nothing to do. There are four channels a knock-on travels through, and you can check each one by name.

Expectation is the one people miss and the most expensive, because it is the hardest to take back.

3. Commit

Three checks decide how you push. They do not decide whether.

Can you undo it, and what does it cost?

Who can actually take this decision? Name the person. A lever you have identified but cannot authorise is a request, and it belongs in a different plan with a different timeline.

What order do you move in? Most change programmes fail because everything moved at once and afterwards nobody could tell what worked. Start with the change that gives you an answer soonest.

4. Measure, and say in advance when you will stop

You cannot measure trust, brand or morale. You measure a stand-in for it.

For each change write five things: what we changed, what we watch, which way it should move, how often we check, and the number that counts as working.

A stand-in measure is worth having only if it passes three tests.

• It moves early. Anything that moves after the thing you care about is a report card.
• It survives being gamed. If your team chased this number directly, would you still get what you wanted? If not, pair it with a second number that guards the obvious abuse.
• It has a threshold attached. A measure with no number next to it is decoration.

Then the step almost everyone skips. Before you start, write down the result that will make you stop. For example: if minutes to serve have not fallen below eight by week six, we reverse the change.

Changes rarely fail loudly. They get quietly defended past the point of evidence, because by then somebody’s name is on them.

Then close the loop. A change that works alters the situation, so the old options are spent and new ones exist. Go back to define.

Framework B. Absorb: change that arrives

Four moves, in order: classify, trace back, choose, watch.

1. Classify the force

Two questions settle most of your response. Did it arrive fast or slowly, and will it pass or is it permanent?

The bottom right box is the reason to do this sorting at all. Fast trouble gets attention on its own. Slow permanent trouble does not, because on no single day is it urgent.

The one operating habit worth building from this is a standing review of what sits in that box.

Where to look for forces: politics, the economy, society, technology, the environment, and the law. Use it as a checklist to scan with. Mark each one as helping, neutral or hurting, then throw away the neutrals. Six paragraphs of commentary, one per heading, is theatre.

2. Trace back to something you can touch

Keep asking why, one layer at a time. What you see, then how people are behaving, then what they are being rewarded for, then how the system is built.

Stop at the first layer you can act on. Going further is satisfying and useless. “It is the macro environment” is a true statement and a dead end.

That layer is, by definition, a lever you hold, which is your way back into Push.

3. Choose your posture

Four options. Pick one deliberately. The failure mode is sliding into the first by default and calling it a decision.

Turning it around is the one to look for. A rule that raises everyone’s costs favours whoever adapts first. A shortage hurts your competitor more than it hurts you if your costs are lower. When you turn a force around you have moved from Absorb into Push, which is what the whole system is for.

Then find the force that matters most. Of everything acting on you, one usually explains most of the movement. Find it before you allocate a rupee of attention.

And name the force pushing the other way. Every situation is a balance of two opposing pressures. You change the outcome by weakening the one working against you or by strengthening the one working for you.

Weakening is almost always cheaper. Most change programmes attempt only the second, which raises the pressure on both sides and moves nothing.

4. Watch for the next one

Same discipline as the measure step, turned around. There you asked whether it worked. Here you ask how early you can see the next one coming.

For each force that matters, write four things: the force, the early sign, the number that triggers action, and what you will do.

Decide that response in advance. Under real pressure nobody designs a good response. People execute what was already written down, or they improvise badly. Write it while you are calm.

How the two halves connect

Without these three joins this is two lists. With them it is one system.

Tracing back finds you a lever. In Absorb you keep asking why until you reach a layer you can act on, and that layer is a lever. Tracing back is how you discover you had agency all along.

Turning it around is a formal handover. The fourth posture in Absorb is the moment you stop receiving the change and start driving it.

Rehearsing from the other side stops bad pushes. Before you commit to any change, sit in the other party’s chair and run Absorb on what you are about to do to them. Ask how they will classify it, which of the four postures they will pick, and how long until they have neutralised it.

If the answer is that they adapt within a month and you are back where you started, the change was never going to work. You found that out before spending anything.

This is also the honest answer to “how do I know my change is good?” You do not. You check who pays for it, and you check what they will do about it.

The whole thing on one page

Start here. Do I hold the lever? Yes, run Push. No, run Absorb. Partly, split it and run both.

The three bridges: tracing back hands you a lever, turning it around moves you from the right column to the left, and rehearsing from the other side tests your push before you spend on it.

Context is just somebody else’s decision. The aim is to spend more of your life in the left column.

The short version

• Every change is either one you are making or one being made to you, and the two need opposite methods.
• If you are making it, the hard work is naming who pays and what breaks.
• If it is happening to you, the hard work is telling a passing shock apart from a slow permanent shift, and catching the second one.
• Whatever is happening to you was somebody’s decision. The goal is to be the one deciding more often.

The bare shape, with no details in it

Every case this framework handles has the same three pieces. A sells something to B. Something outside them both changes: a cost, a rule, a new competitor. Fill in those three blanks and the rest of the framework runs itself.

The chain runs in one direction. An outside change lands on A. A absorbs it and decides something. A’s decision lands on B. B absorbs it and decides something of their own, and so on down the line.

Read it in four beats.

1. The outside change lands on A, who did not choose it. A runs Absorb: is it fast or slow, will it pass or is it permanent, keep asking why until something A controls turns up, then pick a posture and set an early sign to watch.
2. Whatever posture A picks becomes a change A now makes. Almost always, part of the cost travels to B.
3. So A runs Push: name what moves and who pays for it, trace how it travels and what it knocks over, check whether it can be undone and who can authorise it, then attach a measure and a written stop rule.
4. Before committing, A sits in B’s chair and runs Absorb from there. If B’s cheapest posture is to walk away, A’s change fails, and A learns that for nothing.

The same picture explains the claim this document rests on. The outside change that hit A was somebody’s decision further up the chain, and A’s decision is now the outside change that hits B. Nobody in the picture is only pushing or only absorbing. You are always somewhere in the middle, and the question is only which end you spend more of your time on.

The canteen below is this picture with the blanks filled in: A is the owner, B is the students, and the outside change is the price of milk.

Worked example: the college canteen

The canteen sells tea at ₹10 a cup. Milk prices go up. The owner is thinking about charging ₹15.

That is two problems, and they get routed separately.

The part he does not control: milk prices

Classify. The rise came fast. Will it pass or is it permanent? A seasonal dip in supply will pass. A change in procurement rules will not. This one judgement decides everything after it, and it is a judgement rather than a calculation, so say which way you are assuming and why.

Assume fast and permanent. Steady it this month, redesign the menu this term.

Trace back. Margin is down, because milk costs more, because tea is what he sells most of, because tea was priced cheap to pull students in. The last layer is a pricing decision he made himself, and that is his way out.

Choose. Taking the hit does not survive a permanent force, because savings run out. Adapt fits: let black tea, samosas and cold drinks carry more of the revenue. There is also a turn-it-around available. If milk costs hurt every canteen and every stall outside equally, whoever changes the menu first has a cost advantage for as long as the others take to follow.

Watch. Early sign: milk cost as a share of the day’s takings, checked weekly. Trigger: if it crosses a set level, the menu change moves up. Decided now, while nothing is urgent.

The part he does control: the price of tea

Define. Move tea from ₹10 to ₹15, for the owner, paid for by students, from the first of next month. Written that way, the trade is visible. Written as “revise pricing”, nobody has to look at it.

Then apply the rule about good and bad. Good for the owner’s margin this month. Bad for students on every horizon. Bad for the canteen next term if students drift to the stall outside and stay there.

Trace. The path is direct, because he sets the price himself. That is the reliable kind. Now the four channels:

Expectation is the killer here. A price rise is close to irreversible, and it appears in nobody’s arithmetic.

Commit. Cheap to do and hard to undo, which puts it in the think-then-do box. A better shaped version: raise the price only at the evening counter, where the outside stall is shut, so the exposure is bounded and he learns something.

Measure. Watch cups sold per day. The month’s takings tell him afterwards, and cups tell him in time to reverse. Stop rule, written today: if evening cups fall more than 20% against the morning counter by week three, the price goes back.

Rehearse from the other side. Sit in the student’s seat and run Absorb. To a student this is fast and permanent, and the obvious posture is exit, to the stall outside. If exit is that easy, the change fails, and he knows it before printing a new board.

What is borrowed, and what is new

Borrowed, and used as technique rather than structure

What this framework adds

1. Routing before analysis. The first question is “do I hold the lever”, because the answer sends you down two different methods. Most change models work the same way in either direction, which is why they read as checklists.
2. Good and bad as a three-part judgement: the change, whose seat, what horizon. This turns a vague argument into a stated trade-off.
3. Four named channels in place of “knock-on effects”: capacity, incentive, expectation, response. A butterfly’s wings are vivid and give you nothing to do. Four named channels give you something to check.
4. Expectation named as a channel of its own. It is the most expensive knock-on and the least reversible, and it appears in no standard framework.
5. A written stop rule fixed before you start. Evidence rarely kills a change, because by the time it arrives somebody owns the change.
6. The slow-and-permanent box as a standing agenda item. Everything else on that grid gets attention naturally.
7. Rehearsing from the other side: running the second framework on yourself, from the other party’s chair, before you commit.
8. The claim that ties it together. Context is just somebody else’s decision, so the two halves are one system seen from opposite ends.

A framework earns its name by making a claim about how the world works. Point 8 is this one’s claim.

Where this framework does not help

It assumes you can tell fast from slow, and passing from permanent. Often you cannot, and you find out afterwards. The grid forces you to commit to a view, which is useful, and the view can still be wrong. The honest use is to state the assumption out loud and set a trigger that tells you when it was wrong.

It says almost nothing about people: resistance, politics, who loses face, who has to be persuaded. The “who can actually decide” check touches this and goes no further. For a change that depends mainly on persuading a workforce, this tells you what to do and leaves out how to get anyone to agree.

It assumes measurement is possible. Some things matter and resist measurement for years, and culture and reputation are the obvious cases. Applied strictly, this would talk you out of changes worth making, because no early sign exists yet.

Nothing here helps you invent the option. It tests changes you have already thought of. Coming up with the good idea is a separate problem.

It is built for one organisation’s point of view. With several parties each pushing and absorbing at once, you have to run it once per party, and it says nothing about how those runs interact.

One last limit, and it is the biggest. If every situation you look at routes to Absorb, that is worth examining. Sometimes it is true. More often the levers sit a level above you, and the real change to make is about who gets to decide.