Guides · Running a lab

Moving a lab off paper without losing the back catalogue

Every lab that switches to an electronic lab notebook hits the same wall: the twelve years of paper already on the shelf. This is what to do with it, what not to bother scanning, and how to avoid the month where everyone keeps the record twice.

Reading time
11 min
Licence
Free to copy into your own lab documentation. No signup.

The shelf of hardback notebooks is not the problem people think it is. The problem is that most labs try to deal with it first, discover it will take four months, and quietly abandon the whole move.

The labs that succeed do the opposite. They start with the new work and leave the shelf alone — then go back for the small fraction of it that anyone will ever open again. What follows is that sequence.

Who this is for

A PI or lab manager who has decided to move, or is close to deciding. It assumes you have paper notebooks, at least one person who is sceptical, and no budget for a scanning service. Most of it applies whichever electronic lab notebook you pick, including ones that are not us.

Do not start with the shelf

The instinct is to digitise the archive first, so that everything lives in one place from day one. It is the right instinct and it is the reason most of these moves fail.

Digitising twelve years of paper is a project with no deadline, no owner, and no visible reward. It gets assigned to whoever is least busy, which means a summer student, which means it stops in September. Meanwhile nobody has started using the new notebook, because the migration "isn't finished."

Start the new record on a Monday. Deal with the shelf later, and only the parts of it anyone will open.

The new work is where the value is. An entry written today, with the file attached, is worth more than a scan of an entry from 2019 that nobody will read.

The month where people keep two records

This is the documented failure mode, and it is worth naming because it is so predictable. Paper by day, typed up in the evening. It is twice the work for no benefit, everyone resents it, and within five weeks the typed version stops.

It happens when the switch is announced for the whole group at once. The fix is unglamorous: one person, one study, no announcement.

Pick someone who is starting something new anyway — a student beginning a project, or a postdoc opening a new line. They keep the new study in the notebook only. Everyone else carries on exactly as before. Nothing is mandated, so nothing is resented, and in six weeks there is either a working example or an honest reason it did not fit.

Mandating it across a group on day one reliably produces thin, box-ticking record-keeping. That is worse than paper, because it looks like a record and isn't one.

What is actually worth scanning

Once new work is running, the shelf becomes a much smaller question. Almost none of it needs to be searchable. Sort it into three piles.

1. Anything behind a published figure

This is the pile that matters and it is usually thin — a few dozen pages across a decade. If a reviewer or a co-author asks about a figure in three years, this is what you will be looking for. Scan it, and name each file with the paper and the figure number.

2. Anything a current project still depends on

Protocols in active use, the standard curve everyone still refers to, the optimisation that settled a method. Do not scan these — retype them. A scan of a protocol is a photograph of a protocol; nobody can edit it, correct it, or follow it on a phone at the bench. This is the one place retyping is worth the hour.

3. Everything else

Leave it on the shelf. Label the shelf. Write one page in the notebook saying which years are in which books and where the shelf is. That page takes ten minutes and does ninety per cent of the work of a full archive.

Why we care about this

woodle keeps a result attached to the run and the original file behind it, so the connection you are trying to preserve on paper is kept for you on the new work. It also imports folders, instrument files and photographed pages on the free plan, so pile one can go in as-is.

See how import works →

Photographing pages is fine

A phone camera is a scanner. Modern phone cameras deskew, crop and sharpen a page better than the flatbed in the corridor, and they do it in two seconds rather than forty.

Two rules make the difference between a useful pile of images and a useless one. Name the file as you go — the notebook, the date range, the project — because a folder of IMG_4471.jpg is not an archive. And photograph the spine and the first page of each book, so there is always a way back to the physical object.

Resist the urge to buy a document scanner for this. It will be used for one week.

What to do about signatures and dates

In labs that countersign notebook pages, this is usually the sticking point, and it deserves a straight answer rather than reassurance.

A signed, dated, bound paper page is a strong record precisely because it is hard to alter without it showing. Any electronic replacement has to earn that back through version history — the record has to be able to show that an entry existed before a conclusion was drawn, and that later corrections are visible as corrections.

So ask directly whether the system you are choosing keeps earlier versions when an entry is edited, and whether that history can be switched off or cleared by an administrator. If it can be cleared, it is not doing the job the countersignature was doing. That question is worth asking before the trial, not after.

The checklist

Work down it. Tick as you go — this list remembers your progress while the page is open.

Moving off paper 0 / 10
Pick one person and one new studySomeone starting something anyway. No group announcement, no mandate.
Agree a review date six weeks outDecide in advance what "this worked" would look like, so the answer isn't a vibe.
Ask the version-history question before you commitAre earlier versions kept, and can an administrator clear them?
Ask what the export looks likeAnd whether it opens without their software. Ask before signing.
Write the one-page shelf indexWhich years are in which books, and where the shelf is. Ten minutes.
Identify every page behind a published figureThe pile that actually matters. Usually much smaller than you fear.
Photograph that pile, naming files as you goPaper, date range, project. Never IMG_4471.jpg.
Retype protocols still in active useDo not scan these. A photograph of a protocol cannot be followed or corrected.
Decide who is allowed to say "not yet"Someone must be able to stop the rollout without it being a defeat.
Keep the paperDo not throw out a notebook because it has been photographed. Storage is cheaper than regret.

What good looks like six months in

Not "everything is digitised." The test is narrower and more useful: a new person can find the evidence behind a recent result without asking anyone, and nobody in the group is keeping the record twice.

The shelf is still there. It should be. What has changed is that it is no longer the only place the lab's memory lives, and that the last two years of it are attached to the files that produced them.

Take it with you

This checklist is free to copy, print, or paste into your lab's own documentation. No attribution needed and no signup — if it's useful, use it.