Splitting One Person Across Two Products Without Halving Their Output

You put someone on two products at fifty percent each and expect two half sized outputs. You get less than that. The gap is not laziness and it is not bad planning. It is the cost of reloading a head full of one product's details every time the person moves to the other one.

The short answer: split people by days, not by hours, keep the smaller side above thirty percent or drop it entirely, and give each side one owner who is not the split person.

The reload is the real cost

Work is not just the typing. Before someone can do anything useful on a product they have to remember where they left off, what changed while they were away, who is waiting on them, and which of the four half finished things is now urgent.

On a single product that reload happens once a morning and it is cheap, because yesterday is still fresh. On two products it happens every time you cross over, and the memory is a day or two stale. The person spends the first stretch of every block reading their own notes.

Nobody puts that on a timesheet. It shows up instead as work that keeps slipping by a day, and as a person who feels busy and looks unproductive.

Day blocks beat hour blocks

The most common split is the worst one: mornings on product A, afternoons on product B. It sounds fair. It means two reloads a day, five days a week.

Try whole days instead. Monday and Tuesday on one, Wednesday to Friday on the other. Two reloads a week rather than ten. The person gets long enough on each side to finish something rather than just start it.

If the work allows it, weeks are better still, though weeks make people harder to reach and can leave the quiet product feeling abandoned. Days are the practical middle.

Some rules that hold up in most teams:

• Put the harder product on the longer block. Deep work needs runway.
• Keep the block boundary on a real day boundary, not lunchtime. In the UAE, the Friday half day is one boundary already fixed for you.
• Do not schedule meetings for product A on a product B day. That is a reload with none of the work.
• Let the person finish a thought. A block that ends mid task guarantees a slow start next time.

Ratios that never work

Anything below about a fifth is not a split, it is an interruption. The person will not get past the reload before the block ends, and the small side will feel like homework. If a product truly needs one day a fortnight from someone, it is better handled as a named list of specific tasks than as a standing commitment.

Three way splits fail almost every time. Two products is already a stretch. Three means the person is always behind on two of them and never has a place where they are the person who knows.

Even fifty fifty is harder than it sounds, because in a fifty fifty split there is no home. Nobody expects the person to know the whole picture on either side, so nobody asks, and the person stops keeping the whole picture. Sixty forty is usually kinder. It says which one is home.

Give each side an owner who is not the split person

The quiet failure of splitting is that decisions pile up on the shared person's off days. Someone needs an answer on Tuesday, the person is on the other product, and the question waits.

Fix that by naming an owner on each side who can decide without them. The split person does the work; the owner keeps the thread moving. If neither side has an owner, the split person becomes a bottleneck in two places at once, which is worse than being fully committed to one. Across a time zone gap, as anyone working between Kerala and the UAE will know, the wait gets longer again.

What a shared record helps with

Most of the reload cost is looking things up. Where did we land on that, who is out this week, which decision is still open. When those answers live in three systems the reload takes longer than the work.

SicherOne puts project management and HR on one set of records, so a person coming back onto a product can see the board, who is on leave, and what moved, without opening three tools and asking two people. AI agents work off those same records and can summarise what changed, though a human approves anything an agent produces before it goes anywhere.

That does not remove the cost of switching. Nothing does. It makes the first hour of the block shorter, which over a month is most of a working day back.

The honest version

A split is a compromise you accept because you cannot hire yet. Say that out loud to the person. Tell them which side is home, which side is the smaller commitment, and who decides when they are away. Splits go wrong quietly when everyone pretends the person is fully available to both, in the same way a plan goes wrong when it ignores the weeks around Onam and Eid.