Reading the Company From Outside

What to read

The product pages. Not to learn the product — to learn the words. Every company has eight or ten terms it uses constantly and defines nowhere, and half of them appear on the public site. Walking into your first meeting already knowing what the four products are called is worth more than it sounds.

The last year of announcements. Blog, newsroom, release notes, whatever exists. This tells you what the company thinks it has been doing, which is not the same as what it has been doing, and the gap is informative later.

Anything public they have written for engineers or users. Open repositories, API documentation, a help centre. The tone of internal documentation is usually the tone of the public kind.

Job advertisements, including ones not for your role. Twenty open roles in one team and none in another tells you where the money and the attention are going.

What you are looking for

Not facts. Four things that are hard to ask about directly.

The third row is the most useful and the least obvious. A product with a page, a price and no mention anywhere in eighteen months is a product in maintenance, and if you have been hired onto it that is worth knowing before your first week rather than during your sixth.

Keep it to an hour

The failure mode here is doing too much.

Reading for a whole weekend produces a head full of detail you cannot place, and it is detail you will acquire anyway in the first fortnight. Worse, it produces the feeling that you ought to already understand things you cannot possibly understand yet, which is the single most common source of unnecessary misery in a first month.

One hour, a page of notes, and stop.

What to write down

Three lists, short.

Words you did not understand. These become your first questions and they are good ones — nobody minds explaining their own vocabulary, and asking shows you have read something.

Things that surprised you. Keep these especially carefully. In six weeks they will have stopped being surprising, and the record is the only way to get them back.

Things that do not add up. A product with no news. Two pages that say different things. A team whose work you cannot find any trace of. Do not raise these in week one — you will be wrong about most of them — but note them, and look again in month two.

What this is not

It is not research for an interview. That is over.

It is also not a way of forming a view about the company before you arrive. You cannot form a useful one from outside, and arriving with conclusions is worse than arriving with none: you will spend your first month finding evidence for them rather than looking at what is actually there.

The purpose is narrower and entirely practical. On your first morning somebody is going to use four acronyms in one sentence, and the difference between recognising two of them and none of them is the difference between following the meeting and watching it.

The thing you cannot read from outside

Worth naming, so you do not go looking for it.

Nothing public tells you what the place is like to work in. Not the careers page, not the announcements, and not the reviews on employment sites, which are written almost entirely by people who left unhappily or were asked to post.

The tone of the public documentation is a weak signal and it is the only one: a company whose user-facing writing is careful and plain is usually not a company whose internal writing is chaotic.

Everything else about the texture of the place — whether people answer each other, whether decisions get made, whether anybody is happy — is discoverable only from inside, and mostly in the second month rather than the first.

So read for vocabulary and shape, and hold the rest open. Arriving with a view formed from outside is worse than arriving with none, because you will spend the first month confirming it.