Tracking backlinks in a spreadsheet: how far it goes
A spreadsheet is the right place to start, and it stays useful far longer than tool vendors like to admit. The problem is not that it gets slow, or that you run out of rows. It is that a spreadsheet records what you were told, and a link is a fact about a page right now.
What the spreadsheet does better than any tool
Everything about the relationship. Who you contacted, when, what was agreed, what you offered, which article it went into, who to write to if it disappears, what it cost you, and the note that says "he asked to be kept posted".
No monitoring tool holds that, because none of it lives on the page. A crawler can tell you a link vanished; it cannot tell you that the person who placed it owes you an answer, or that you agreed to leave it alone until September. If you drop the spreadsheet when you adopt a tool, you lose the half of your link profile that is actually about people.
The columns that hold, and the ones that rot
| Column | Who keeps it current | How long it stays true |
|---|---|---|
| Source page URL | Nobody, it does not change | Indefinitely |
| Contact, date, what was agreed | You, once | Indefinitely |
| Cost, consideration, end date | You, once | Indefinitely |
| Judgement: is this link worth effort | You, once | Years |
| Status: live, lost, nofollow | Nobody | A few weeks |
| Anchor text recorded | Nobody | A few months |
| Date last checked | You, on every pass | The honest column |
The split is clean: anything describing a decision stays true, anything describing a state goes stale. A spreadsheet excels at the first half and has no way of maintaining the second.
The one column that lies
Every backlink spreadsheet has a status column. Live, lost, nofollow.
That column is true on the day you type it and starts decaying immediately. Three months later you are looking at a green cell that describes a page you have not opened since spring. The spreadsheet does not know that; worse, it does not look like it does not know. A stale cell and a fresh one are the same shade of green.
That is the whole limit, and it has nothing to do with size. Forty rows you last verified in March are less trustworthy than four hundred verified last night.
The minimal fix, if you change nothing else about your method: add a "checked on" column, and colour the row by its age rather than by its status. You will then be looking at what you actually know.
When manual checking genuinely holds up
- Under about twenty links you care about, and a habit of opening them.
- Links you cannot influence anyway: a press mention, a forum post, an archive. Knowing sooner would not change what you do.
- A campaign that is finished. Nothing new is coming in; you are curious, not responsible.
- A very stable profile, where links come from sites that have been around for ten years and that nobody rebuilds.
If you are in one of those, a monthly hour with the spreadsheet open is a perfectly rational choice, and cheaper than any subscription. Nobody should talk you out of it.
Where it breaks, concretely
It breaks when someone else depends on the answer. A client asks whether the links you reported in March are still up. You either spend an afternoon opening pages, or you answer from a column you no longer trust.
It also breaks silently on the changes that are not removals: a link left in
place but switched to nofollow, an anchor rewritten during an edit, a source page
set to noindex. Those never announce themselves, and no amount of discipline
makes you spot them by eye, because there is nothing to see:
they are the changes only a comparison over time reveals.
And it breaks at the most expensive moment: during a redesign, yours or a partner's, when several dozen links change state in the same week. That is exactly where a quarterly spreadsheet arrives too late.
The template that works, if you are starting from scratch
Seven columns are enough, and the order matters because it tells the life of the link:
- Source page: the exact URL, never the site name alone.
- Your target page: the one that has to keep answering.
- Contact and date: the person, not the company.
- Origin: editorial, guest post, exchange, paid. This column prevents unpleasant conversations two years later.
- Judgement: worth an effort, or not. Written once, deliberately, using the criteria that make a link good.
- Checked on: the only honest state column.
- Notes: everything else, free text. It is the most valuable column and the one no tool will ever produce.
What does not deserve a column: authority scores copied by hand, which will be wrong in six months, and toxicity statuses, which serve only to worry you. Disavowing is decided on other grounds.
What a manual check really costs
People rarely do this calculation, and it settles the question better than any
argument. Checking a link seriously means opening the page, searching the source
for your domain, reading the rel attribute, recording the anchor and noting the
date: about ninety seconds when all goes well, more if the page is heavy or the
link has moved.
| Number of links | One full pass | If you do it monthly |
|---|---|---|
| 20 | 30 minutes | 6 hours a year |
| 60 | 1 hour 30 | 18 hours a year |
| 200 | 5 hours | 60 hours a year |
What that table shows is not that manual checking becomes impossible: it is that it becomes irregular. Nobody blocks out five hours a month to open pages, so the pass slips by a month, then two, and the status column ages with nothing to flag it. The tipping point is not where the work becomes too long, it is where it stops being done.
Three mistakes found in almost every spreadsheet
Recording the site name instead of the exact URL. "Someone's blog" cannot be verified. Six months later nobody knows which page the link was on.
One row for three links from the same site. They do not live the same life: one is in an article, another in a resource page, the third in a footer. One row per link, always.
Deleting the rows of lost links. The memory of your outreach leaves with them. An archived row, with the date it disappeared and the reason, beats a tidy file: it is what stops you asking in 2027 for a link the same site has already removed from you twice.
The arrangement that works
Keep the spreadsheet for the relationship and for the judgement. Let something check the pages on a schedule and tell you when a state changes. The spreadsheet then holds only things that are true because you decided them, which is exactly what a spreadsheet is good at.
This split has a pleasant side effect: it makes the spreadsheet shorter. The state columns disappear, the decision columns stay, and the file becomes readable again.
Related: why links disappear in the first place, which explains why the freshness of a check matters more than its thoroughness.
Frequently asked questions
Is a spreadsheet enough to track backlinks?
For the relationship side, yes, and often better than a tool. For the state of each link today, no: a spreadsheet records what you were told, not what the page currently contains.
At what point does it stop working?
Not at a number of rows. It stops when the "status" column becomes older than your memory of filling it in, which usually happens well before you notice.
Should I throw the spreadsheet away?
No. Keep it for what it is good at, who, when, what was agreed, who to contact. Let something else answer whether the link is still there.
Never lose a backlink again
Add your sites and links, and let Expansel watch them for you.
Start free