Backlink Profile Worksheet: Columns and First Findings

A backlink profile is simply the working name for everything a site receives in links from other sites: from whom, to which page, with what text and with which attribute. That is the whole definition, and we make no claim that Google scores a profile under that name. Instead of explaining the term at length, this article shows how to turn an export you already have into a backlink profile worksheet: split it into columns, group sources by domain, read the target and anchor spread, and convert first findings into an action list. Every record in the tables is hypothetical, matches no real site and promises no outcome.

A Backlink Profile Worksheet: From Definition to Table

The wider logic of reading a profile and the checker tools around it is covered in reading a backlink profile with link checkers. We enter from a different door: you have an export file and want to make it workable, row by row.

Picture the owner of a hypothetical hiking-gear shop who receives a spreadsheet of several hundred link rows from an agency. The data is accurate and complete, yet it answers nothing, because no question was written down first. A worksheet is the step that turns questions into columns.

A method, not a tool

A worksheet lives in any spreadsheet program and replaces no ready-made report. It rearranges raw data around your questions: which page attracted the most links, which source forms a cluster, whether anchors concentrate in an unusual way. A single table should be able to answer those.

Three decisions before you start

Write down whose profile is being read, which sources the data will come from and which decision the worksheet serves. Skip the third and you get a handsome table that leads nowhere. The decision might be choosing a target page for new links, or it might be flagging odd sources for review.

Exporting Profile Data and Keeping a Raw Copy

Every worksheet starts from an export. For the basic idea of what a link is, the introduction to backlinks is a good companion; this section covers only how the data is obtained and protected.

Leaving the raw file untouched

Save the export exactly as it arrives, put the date and source in its filename, then open a working copy as a separate file. As long as the raw file stays unchanged you can start over after a mistake. In the working copy, deleting, sorting and merging columns are all fair game.

Combining several sources in one table

Profile data often arrives from more than one place: a link monitoring tool, a search console report, a manual list. Keep a column that records the origin of each row, and when one link appears in two sources, write both in that column rather than duplicating the row. Counts from tools are tool-specific, not Google values, so do not try to make totals match across sources.

Spotting a capped export

Many tools and reports cap what they export. Do not interpret counts before you know whether the file is complete or a sample. A note in the top row, "complete" or "sample, limit unknown", lowers the chance of misreading.

Columns: Source, Target, Anchor, Label and Date

The columns decide what the reading can rest on. Too few leave questions unanswered; too many make the sheet hard to maintain.

The seven core columns

Source URL, source domain, target URL, anchor text, link attribute, first-seen date and data origin. Each answers one question: from whom, from which page, to where, with what text, with what attribute, since when and according to what.

Allowing "unknown" for the attribute

Not every tool shows link attributes, and the Search Console Links report does not show whether a link is nofollow. When the information is missing, write "unknown" and do not guess. Google's guidance on marking outbound links assigns sponsored to ads and paid placements, ugc to user-generated content and nofollow to everything the others do not cover, so keep those three values plus "none" and "unknown" as a fixed pick list.

Two extra columns: cluster and first note

During the first read, give each row a cluster name such as "directory entry", "news article", "forum" or "unclear", plus a one-sentence note. These two columns save you from re-sorting the sheet during grouping.

Grouping Sources by Domain

Grouping reduces hundreds of rows to a handful of meaningful clusters. A fuller look at how varied your link sources should be is in the article on diversifying a backlink profile; here we only deal with how to group.

The root domain rule

The Search Console Links report groups data by root domain: subdomains and subdirectories merge, while top-level domains stay separate. If you adopt the same rule in your worksheet, comparing the two becomes easier. If you change the rule, write it in the top row, otherwise grouped counts will not match between sources.

Many links from one domain

When one domain produces dozens of rows, ask two separate questions: do the links come from different pages, or are they repeats of a single template? Many links from one structure are often the echo of one relationship, and treating the group as a single row is more realistic.

The long tail of one-link domains

The long tail, domains that give only one link, is information too. Count it as its own cluster and label it by source type; it gives a quick sense of the overall shape of the profile.

Reading the Target Page Distribution

Links do not spread evenly over a site. Seeing the distribution shows which pages draw outside interest and which draw none.

A pivot by target page

In the working copy, build a summary by target URL: row count per target, number of distinct source domains and the oldest and newest first-seen dates. Counts are tool-specific, but comparisons inside one dataset still tell you something.

Home page versus inner pages

If most links point to the home page, outside interest in service or product pages may be limited. That is an observation, not an error, and it feeds the plan for new links. Mark with the business team which pages carry commercial weight. A commercially important page with almost no outside interest may be the first target for acquisition, although that does not mean a link to it will produce a result; it only shows a gap in the profile.

Also look at the variety of sources behind each target. If all of them are directory entries, variety is low; a mix of news articles, forum replies, blog mentions and directories looks more balanced. Record this as a reading note rather than a verdict.

Whether the target still exists

Some links point to pages that no longer exist or now redirect. Note the status code in a separate column; broken targets are often the cheapest finding to repair.

Reading the Anchor and Label Distribution

Anchor texts show how a profile describes itself. Planning anchor text for a link profile covers targeting logic; in the worksheet we only look at the spread.

Sorting anchors into types

Assign each anchor a type: brand name, bare URL, generic phrase ("click here"), topic phrase and product or service phrase. Summarise the spread in one sentence. Google's crawlable links page wants anchor text that describes the destination, stays reasonably short and fits the page, and it warns that keyword stuffing is a spam policy violation. An unusual concentration in the spread is a flag for review from that angle.

The attribute spread

Count the attribute column too: how many rows show sponsored, ugc, nofollow, none or unknown? A high share of unknown is a data gap, and it should make you more careful with every finding.

Empty anchors and image links

Links given through an image can show an empty anchor, and the Search Console report also labels links without text as "(empty)". Do not count them as problems; mark them as their own type. Likewise, plain-text brand mentions without a link do not belong in this table, since they are not links.

First Findings: Suspicious, Unexpected and Healthy Clusters

First findings are the patterns you notice, sorted into three boxes. For the limits of tool signals such as spam scores, see the explanation of spam score and what lowers it; those scores are tool-specific, not Google values.

Suspicious: patterns that need review

Many links from one template, directory entries unrelated to the topic, and placements that look paid but are unqualified belong here. The box is called "suspicious", not "bad", and every row gets a sentence explaining why it was flagged.

Unexpected: unexplained but possibly harmless

"Unexpected" is the box you cannot fill without first knowing what you expected. A link that arrived as a quotation in a blog post or a forum reply may sit here. The goal is a record, not a decision: why is it there, who placed it, is it still live?

Healthy: clusters to protect and repeat

Links in topical publications, with descriptive anchors and correct qualification, go here. Write down what healthy clusters share: which source type, which target, which context. New acquisition looks for placements that match that description.

Cross-Reading With the Search Console Links Report

Comparing your export with Search Console data is a cheap way to find gaps. Query and page performance is a separate topic, handled in the article on Search Console performance by query and page; here we look only at the Links report.

What the report shows and what it leaves out

The report presents external links as top linking sites, top linked pages and top linking text. It merges duplicate links from the same source to the same target, caps tables at 1,000 rows and, as its own help page says, is not a comprehensive list but a sample; removed links or pages can still show up. It does not show whether a link is nofollow.

A cross-reading table

Sources that appear in the report but not in your tool data, or the reverse, carry information. Label them "report only" and "tool only". The gap may simply come from different coverage, so write it as a follow-up question, not a verdict.

Why report and tool counts disagree

Disagreement is normal: grouping rules, data coverage and refresh dates differ. Rather than forcing the numbers to match, state in one sentence at the top of the worksheet what each source measures.

Turning Findings Into an Action List

A finding that never becomes an action leaves the worksheet in the archive. Start with the smallest effective action. The disavow process has its own explanation in how the backlink disavow process works; at this stage we only decide which finding goes to which action.

Five action types

Five are enough: no action, monitor, fix (for example redirect a broken target), contact the publisher, collect more information. Assign every finding one of the five. Bigger steps such as removal requests or a disavow file are considered only when smaller ones have not worked.

Three questions for priority

Is the finding about risk or opportunity, how much effort does a fix take, and what is affected if it stays? Low effort and high importance come first. Record results in a priority sheet with an owner and a review date.

Linking new acquisition to the findings

Opportunity findings can feed new link acquisition. Akua offers no profile analysis or auditing tool; you keep the worksheet. If you do place an order, it goes like this: you pick a package and complete payment, the seller approves the job and sends a delivery notice when it is finished, and you can raise a dispute if the outcome does not match the listing. The acquired link enters the worksheet as a new row at the next export.

Storing and Versioning the Worksheet

A worksheet is not a document made once and left; it is a record you compare over time.

File names and a version scheme

Put a date and a version number in the filename, for example profile-worksheet-2026-10-v1. Create a new file at each refresh and never delete the old one, so two versions can be read side by side and the movement of clusters becomes visible.

The change note

Add three lines at the top of every version: what changed, which findings closed, which new questions opened. The note lets someone who returns to the sheet after six months regain context quickly. It should also say whether the data source changed, whether a grouping setting was adjusted and which earlier actions are complete; without those answers the difference between two versions cannot be read.

Access and sharing limits

If the worksheet holds client or internal commercial information, restrict access. When sharing, send summary findings instead of raw data; that protects confidentiality and keeps the recipient from fixating on the wrong rows.

Example Template: A Profile Worksheet (Hypothetical)

Every domain, anchor and date in the table is invented and points to no real site. It shows seven rows from the profile of a hypothetical hiking-gear shop, with first-read notes. When an attribute is not known, the cell says "unknown".

Hypothetical first findings: two rows of an off-topic directory entry from one template (review), one broken target (fix the redirect) and one unclear forum link (monitor). When you adapt the sheet, keep these points fixed:

• Never touch the raw file; work on a separate copy.
• Write the grouping rule (root domain) in the top row.
• Add a one-sentence reason to every suspicious row.
• Write "unknown" where data is missing instead of guessing.

At each refresh, follow this order:

1. Store the new export untouched and open a working copy.
2. Compare source and target columns with the previous version.
3. Sort new rows into clusters and write the first notes.
4. Tie the findings to action types, with an owner and a date.

Frequently Asked Questions

What is a backlink profile?

A backlink profile is the full set of links a site receives from other sites: source pages, target pages, anchor texts and link attributes. Here it is a working concept; nothing suggests Google applies a profile score under that name. Reading the profile means putting that set into a table and grouping it to see what it shows.

Which columns belong in a backlink profile worksheet?

The core columns are source URL, source domain, target URL, anchor text, link attribute when known, first-seen date and where the data came from. Add a cluster column and a first-note column. Rather than adding more columns, state which question each one answers, and remove any column with no answer.

Does the Search Console Links report show every link?

No. The report is a sample, not a comprehensive list; tables stop at 1,000 rows, removed links can still appear and duplicate links are merged. It also does not tell you whether a link is nofollow. Use it as a cross-check against data from other sources, not as the master dataset.

Should every suspicious link in the profile be rejected?

No. A suspicious label is a flag for review, not a verdict. Look at the cluster level: a group from one source may matter less than it seems, while a single odd link can matter more. Find out why first, then choose the smallest action that fits. Rejecting everything costs time and discards information.

How often should I refresh the worksheet?

Monthly during periods of active link acquisition, and every three months when things are quiet is often enough. Save each raw export as a new file and keep the old ones. That way you can watch cluster counts and new sources change over time and follow what your action list actually achieved.

How do I prioritize the first findings?

Ask three questions: does the finding concern a risk or an opportunity, how much effort would a fix take, and what is affected if it is left alone? Take low-effort, high-importance items first. In the priority sheet, put an owner and a review date next to each finding.

Conclusion and Next Step

The definition of a profile is short; the real work lies in cutting it into rows and asking questions of it. With an export, columns, root-domain grouping, target and anchor spreads, three-box first findings and an action list in place, you hold a record that can be read again later. It says nothing about rankings, traffic or any other result; it only preserves what you observed and why you acted.

If your worksheet marks a target page as an opportunity and you are considering new links for it, you can look through the products on Akua's backlink packages page. Package contents, delivery terms and the dispute route are stated on each listing; your worksheet decides which action fits.