How to Compare SEO Proposals Side by Side

Three proposals land in your inbox. One asks RM2,800 a month, one RM5,500, one RM9,000. All three promise better rankings, more traffic and a monthly report. And there is no honest way to say which is better value, because none of them describes the same work.

That is the actual problem when you compare SEO proposals. The prices differ because each one is attached to a different thing, written in different words, over a different period. A spreadsheet with three columns and a one-to-five score then measures something real, but not the thing you wanted. It measures which agency writes the most convincing document.

The fix is unglamorous and it works. Before scoring anything, you rebuild all three proposals onto one sheet in one unit. A scorecard only means something after that.

The stakes are ordinary business stakes, not marketing ones. Malaysian MSMEs produced RM652.4 billion in value added during 2024, or 39.5 per cent of national GDP, according to the Department of Statistics Malaysia. Almost none of those owners has a procurement team to normalise a quote for them. Most sign on gut feel and a price.

So this piece runs the normalisation in public. What genuinely differs between proposals, which lines can be compared at all, how to rebuild three documents onto one sheet, and what to test before signing. That sequence is how we read a client’s shortlist at IZI Digital Marketing before recommending anything. First, why the documents refuse to line up on their own.

PART 1 · DIAGNOSE

Why Three SEO Proposals Never Compare Cleanly

IN BRIEFProposals differ in unit, not only in price. One sells deliverables, one sells hours, one sells outcomes — so the same ringgit buys three different promises. Decide which unit you will compare in before reading any of them, then judge each against the same SEO scope.

Agencies are not being slippery when their documents refuse to line up. They are describing genuinely different businesses. A specialist and a full-service shop solve your problem in different orders, and each writes down what it sells. The usual complaint is that scopes are vaguely worded, but vagueness is only the symptom. The cause is that the three documents are not counting in the same currency.

Three mismatches do almost all the damage:

• Different units of sale. One quote promises four articles and twenty backlinks. One promises twenty-five hours. One promises a traffic figure. None of these converts into the others without your assumptions doing the work.
• Different time horizons. A six-month scope and a twelve-month scope at the same monthly fee are not the same purchase. Total cost, not monthly fee, is the honest comparison.
• Different starting assumptions. One agency has already spotted a technical fault and priced fixing it. Another has not looked and priced content instead. The gap is diagnosis, not value.

You cannot remove these differences. You can only convert them into one unit, and the conversion is your job, not theirs. Nobody writes a proposal to be easy to compare against a rival’s.

BENCHMARK BRIEFING 1 OF 4

Which Proposal Lines Can Be Compared At All?

IN BRIEFGoogle’s hiring guidance names eight things an SEO commonly provides. Roughly half arrive in proposals with a countable unit attached; the rest arrive as labels. The half without units is where quotes drift apart, and where a low budget quietly disappears.

Start from a neutral list rather than from any one agency’s menu. Google publishes the service categories an SEO typically covers, so map each proposal onto that list before you read a single price.

PART 2 · DIAGNOSE

Which Words Hide the Same Job Under Different Names?

IN BRIEFHalf the apparent difference between quotes is vocabulary. Rewrite every line into plain description of the actual task, and duplicate work collapses into one row. Do this before pricing anything, the same way you would when shortlisting a digital marketing agency.

Translate each line into what a person will physically do, in one sentence, with no product names. The exercise is dull for about fifteen minutes and then it starts deleting rows.

• “Content strategy and editorial calendar” usually means someone lists article titles and dates. Ask how many titles, and who writes the articles themselves.
• “Authority building” usually means outreach for links. Ask how many placements a month and on what kind of site.
• “Technical health monitoring” usually means a tool runs a scheduled crawl. Ask who reads the output and who fixes what it finds.
• “AI search optimisation” is the newest and vaguest. Ask for the specific deliverable, then check the answer against Google’s own guidance on optimising for generative AI.

Two proposals often turn out to sell nearly identical work once translated. The remaining gap is then small enough to argue about honestly — and small enough that price finally means something.

BENCHMARK BRIEFING 2 OF 4

Where Do Quotes Drift Furthest Apart?

IN BRIEFLines with a natural counting unit stay close across quotes. Lines without one drift, because there is nothing to anchor them. Score each line’s comparability first, then spend your clarification questions on the bottom of the list rather than the top.

The model below rates how easily each common proposal line can be put on a like-for-like footing. Higher means the line already carries a countable unit; lower means you must supply the unit yourself.

PART 3 · DESIGN

How Do You Rebuild Three Proposals Onto One Sheet?

IN BRIEFBuild the sheet from your own scope list, not from any proposal’s table of contents. Every agency then gets scored on the same rows, including the rows they left out. Missing rows are findings, and they are usually the most useful thing the exercise produces.

How to normalise three SEO proposals onto one comparison sheet

The sequence below takes an afternoon and replaces guesswork with arithmetic. Work through it once, in order, before any agency call.

1. Write your rows first. List the scope categories you actually need, in your own words, before opening any proposal. This stops the loudest document from setting the agenda.
2. Drop each proposal into the rows. Copy each line into the row it belongs to. Anything that fits nowhere goes in an “extras” row for later.
3. Convert every line to one unit. Pick either monthly deliverables or monthly hours and convert everything to it. Ask the agency where a conversion is unclear.
4. Mark the blanks. An empty cell means that agency did not price that job. Decide whether you need it, and if so, price it separately.
5. Restate cost over the full term. Multiply monthly fee by contract length, add setup fees, and compare total cost of ownership rather than headline rate.
6. Only now, score. Apply weights to the criteria, fill the scorecard, and read the result before revisiting price.

Anything an agency refuses to convert into a unit is telling you something, politely. Note it and move on.

BENCHMARK BRIEFING 3 OF 4

Which Criteria Deserve the Most Weight?

IN BRIEFWeight scope precision and evidence above price, and treat price as a tie-breaker among the proposals that already passed. The model below splits the criteria into what decides the work and what decides your risk if it goes badly.

Set the weights before you score, and write them down. Weights chosen after seeing the scores are not weights, they are a justification.

PART 4 · DESIGN

Cheapest, Broadest or Deepest: Which Proposal Wins?

IN BRIEFOnce normalised, most shortlists reduce to three shapes: cheapest, broadest and deepest. The right pick depends on whether a diagnosis already exists. Without one, you are buying the diagnosis, and the scope should say so plainly.

The shape matters more than the fee because it decides what you learn. A narrow scope that finishes teaches you whether that lane works. A broad scope that half-finishes everything teaches you nothing you can act on.

BENCHMARK BRIEFING 4 OF 4

How Long Should a Fair Comparison Take?

IN BRIEFAbout thirty days, most of it waiting. The part you control is short: one brief out, one normalisation pass, one shared clarification round. Rushing the brief is what forces the long, messy comparison later.

The timeline below assumes three bidders and one decision-maker. Compress it if you must, but never compress the brief.

PART 5 · DEPLOY

Which Claims Should You Test Before You Sign?

IN BRIEFTest the claims you can check yourself in an hour: past results, guarantees, tool endorsements and AI-visibility advice. Google publishes the checklist, and the same discipline applies when you are reading an agency’s case studies.

Google’s hiring guidance is the closest thing the industry has to a neutral referee, and it was refreshed in June 2026 to cover AI-optimisation services. Four checks come straight out of it:

• Ranking guarantees are disqualifying. Google states plainly that no one can guarantee a number-one ranking, and warns against claims of a special relationship with Google.
• Tool endorsements are false. Google does not evaluate or endorse third-party SEO tools, so any pitch describing a tool as Google-approved fails on a fact, not on taste.
• AI-visibility advice must match the official guidance. Ask the agency to point at the passage its AEO or GEO recommendation follows. If it cannot, the line is a label.
• Audits come before access. Grant read access to Search Console at the audit stage, never write access, and expect realistic estimates rather than promises.

PART 6 · DRIVE

What Should the Winning Proposal Prove by Month Three?

IN BRIEFNot rankings, and not traffic. Month three tests whether the proposal was honest, so check delivery against your normalised sheet and nothing else. Judging early rankings instead is how buyers talk themselves out of a scope problem, as our 2026 review of KL agencies sets out.

Most guidance tells you to look for climbing impressions by day ninety. That is the wrong test this early. Impressions move for reasons no agency controls, and a thin month can still look busy in a chart. Write a delivery test into the contract instead, while you still have leverage. Three checks are enough:

• Delivery matches the normalised sheet. The counts you agreed were shipped, or the shortfall was flagged in the month it happened.
• Pages are indexed and crawlable. New content appears in Search Console coverage, and the technical fixes were verified rather than merely listed.
• Reports arrive on the agreed dates. Late or reshaped reporting in month three is the earliest reliable warning that the account is under-resourced.

If all three hold, keep going and judge outcomes at month six. If any fails, you have a specific conversation to have, with evidence, before the annual renewal removes your options.

THE VERDICT

Normalise First, Score Second

Comparing SEO proposals feels like a judgement problem and is mostly a translation problem. Three documents describe overlapping work in different vocabularies, over different terms, with different assumptions baked in. Score them as written and you reward whoever writes best.

So do the boring part first. Write your own rows, convert every line to one unit, mark the blanks, restate cost over the whole term. Then apply weights that put scope precision and evidence above price, and test the four claims Google tells you to test. The winner usually becomes obvious at that point, and it is frequently not the cheapest or the most polished document.

One last thing worth keeping. The normalised sheet outlives the decision. It becomes the delivery checklist for month three and the negotiating document at renewal, which makes the afternoon it costs you the cheapest work in the whole process.

FAQ

Frequently Asked Questions

1. How do you compare SEO proposals fairly?

Rebuild them onto one sheet of your own rows before scoring anything. It depends on picking a single unit — monthly deliverables or monthly hours — and converting every line into it. Blanks then show which agency simply did not price a job, which is usually the most useful finding.

2. What should be included in an SEO proposal?

A countable unit on every line, plus who does the work and what happens if targets slip. It depends on your starting point, since a site with technical faults needs those priced separately. Google’s hiring guidance lists the service categories worth checking a proposal against.

3. How many SEO proposals should I request?

Three is usually enough, and five is usually too many. It depends on how different your bidders are — three specialists in the same lane teach you less than one specialist, one generalist and one consultancy. More documents mostly add normalisation work, not clarity.

4. Should I pick the cheapest SEO proposal?

Only after it has passed the same scope test as the others. It depends on whether the low fee reflects a narrower, honest scope or the same scope thinly funded. Compare total cost across the full contract term, not the monthly figure, before deciding anything.

5. What should I ask before signing an SEO contract?

Ask who owns the accounts, what the notice period is, and what you keep if you leave. It depends on the contract length, since a twelve-month term with no exit needs stronger evidence upfront. Also ask what the agency would stop doing if month three underperforms.