Do backlinks still matter? Yes — as a tiebreaker

What a link is actually doing in 2026

The founding logic hasn't changed. A link from one site to another is a citation, and citations are the closest thing the open web has to a reputation system. Google confirmed links as a signal decades ago and has never withdrawn that.

What has changed is where in the process links get consulted. Google first has to decide which pages are even plausible answers to a query — a judgement made on meaning, format, coverage and trust. Links contribute little to that stage. Then, among the plausible answers, it has to order them. That's where links do their heaviest work.

Which reframes the whole budget question. Money spent on links is money spent on winning a contest you have to qualify for first. Spend it before you've qualified and it does nothing measurable, which is exactly what most founders describe when they say link building didn't work for them.

Where links genuinely decide the outcome

These are the query types where two pages look equally good to Google and something has to break the tie. Links break it more often than anything else you can influence.

Where links demonstrably don't help

This list is longer than the one above, and it covers most of what a mid-size Indian business actually competes for. If your target queries live here, a link budget is the wrong allocation.

• Local pack results. Proximity, profile completeness and reviews decide the three map slots. A hundred referring domains won't move you closer to the searcher. Covered properly in what local SEO is.
• Long-tail informational queries. Frequently won by whichever page answers the specific question most directly. Sites with almost no external links win these constantly.
• Format mismatch. If Google has decided a query wants a product page and you've published a blog post, links won't promote the blog post. It's not competing in the right class.
• Pages that aren't indexed. No index entry, no ranking, no matter what points at it. Check before you spend.
• Brand and navigational queries. Your own name. If you're not winning these, the problem is entity confusion or a competitor bidding on you, not authority.
• Fresh-intent queries. News and trending topics where publishing speed and publisher authority outrank accumulated link equity.
• Thin or duplicated content. Links to a page Google considers low-quality are wasted, and occasionally worse than wasted.

Velocity, and what natural growth looks like

Link velocity is the rate at which new referring domains appear. There's no safe number, because the right rate depends entirely on what your business is doing — and that's the actual test. Google isn't checking your links per month against a threshold. It's checking whether the pattern has a cause.

A company that launched a product, published original research and got covered in three trade publications should show a spike. That spike has an explanation sitting right next to it. A dormant six-page site that gains 300 referring domains in a fortnight has no explanation, and the pattern is trivially detectable.

• Healthy: steady accumulation that tracks your publishing and PR activity, with occasional spikes tied to something real.
• Suspicious: a flat line for two years, then a vertical climb, then a flat line again. That's a purchase, and it looks like one.
• Also suspicious: perfectly consistent growth — exactly 20 new domains every month for a year. Nature isn't that tidy. Subscriptions are.
• Worth knowing: losing links is normal. Sites redesign, articles get deleted, publications fold. A profile that only ever grows is a profile being manufactured.

What Google's link spam systems actually do

The mental model most people carry is from 2012 — Penguin, manual penalties, sites vanishing overnight. That era ended. Google's current link spam systems, built on SpamBrain, are designed to nullify unnatural links: identify them, discount them to zero, and rank the site as though they never existed.

That's a meaningful change in what failure looks like. You don't get a warning. Traffic doesn't crater. You simply paid ₹2,00,000 for links that contribute nothing, and because there's no visible event, nobody tells you. Plenty of agencies keep invoicing for exactly this, and their reports still show the link count rising.

Manual actions still exist and still arrive with a notice in Search Console. They're rarer, and they usually follow a pattern somebody built on purpose — a large volume of obviously paid, obviously commercial-anchor links pointing at money pages.

Worth saying plainly: negative SEO by pointing bad links at a competitor is largely ineffective for the same reason. Google's systems have to assume any site can be attacked, so nullifying rather than penalising is the only sane design.

Links or content? A ten-minute diagnostic

Before approving any link budget, run this on your five most valuable target queries. It costs nothing and it settles the argument that usually gets settled by whoever is most confident in the meeting.

1. Is the page indexed? Check Search Console. If not, stop here — nothing else matters until it is.
2. What position is it in? Below 30, or not ranking: this is a relevance or format problem. Between 8 and 20: links are a plausible cause.
3. Does your page format match the top five? If they're all category pages and yours is an article, rewrite the page. No link fixes this.
4. Compare page-level referring domains — yours against each of the top five. Not domain-level totals, which are meaningless for this question.
5. If any competitor outranks you with fewer, links aren't your constraint. Something about their page is better and it's usually visible within two minutes of reading it.
6. If all five have substantially more, and your page genuinely matches intent, you've found a real link gap. Now a budget makes sense — and the argument for spending it on internal linking first is in internal linking beats link building.