Most SEO advice is a list of eighty things. Eight of them do almost all the work: match what the searcher actually wants, give each page one job, answer the question in the first sixty words, link your own pages together properly, make it load fast on a phone, write a title somebody wants to click, publish the prices your competitors hide, and earn a handful of real links. The rest is either maintenance or folklore.
SEO has a credibility problem, and it earned it. The field is full of checklists that confuse activity with results, tools that score your page out of a hundred against rules Google never published, and agencies that report on keyword counts because keyword counts are easy to report.
Meanwhile the techniques that actually move a page are unglamorous, mostly free, and have barely changed in a decade. They are just harder to sell than a monthly report with a green circle on it.
What follows is the list that survives contact with real sites. Each one comes with what it is, how to do it, and roughly how much it is worth.
What Changed, and What Did Not
Two real shifts happened in the last few years, and being honest about both is the only way to judge the rest.
AI answers now sit at the top of a lot of searches, which means informational queries lose clicks even when you rank well. Somebody asking "what is a sitemap" often gets the answer without visiting anybody. That is a genuine loss and pretending otherwise helps nobody.
At the same time, commercial searches got more valuable, not less. Nobody hires a plumber, picks a hospital or buys a ₹60,000 website from a summary box. Those searches still end on somebody's page, and that page still has to earn the position.
- What lost value: thin definitional content, listicles that restate common knowledge, anything a two line answer covers completely.
- What gained value: pages with real prices, real numbers, first-hand experience and specifics a summary cannot compress away.
- What never changed: matching intent, one page per job, speed, and links from sites that actually exist.
The Eight Techniques That Still Move Rankings
1. Match the Intent, Not Just the Keyword
This is the technique that makes or breaks more pages than every technical fix combined, and it takes ten minutes.
Before writing anything, search the phrase you want to rank for and look at what Google is already showing. If the first page is full of buying pages, a blog post will not rank there no matter how good it is. If it is full of guides, your service page will not rank there either. Google has already told you what kind of page wins that search. Argue with it and you lose.
Do this check on every page before it is written, not after it fails. It costs nothing and it is the cheapest win in this entire article.
Volume tools are useful at the next step, once you know what kind of page you are writing. Google Keyword Planner will tell you roughly how many people search a phrase, but it will never tell you what they wanted when they searched it. Only the results page tells you that, and it is free to look.
2. Give Every Page Exactly One Job
The most common serious fault we find on Indian business sites is two or three pages chasing the same phrase. The owner thinks they have tripled their chances. What actually happens is that Google picks one, usually not the one you wanted, and all three sit around position twenty.
We audited a site recently with nine pages about the same service. Twenty-two pages covering one topic, forty-four clicks between them in three months. Merging them into one strong page and redirecting the rest is almost always worth more than writing anything new.
- List every page you have and write its one target phrase next to it.
- Any phrase appearing twice is a problem. Pick the stronger page and redirect the other with a 301.
- Check the titles too. Six pages whose titles all end in the same city name are competing whether you meant them to or not.
3. Answer the Question in the First Sixty Words
Most pages open by clearing their throat. Two paragraphs of background before the actual answer appears. Readers leave during that warm-up, and so does the summary box that might have quoted you.
Put the answer at the top. If somebody searched "website development cost in Jaipur", the first line should contain a rupee figure. You can add the caveats underneath, and they will be read, because the person now trusts that you are going to tell them things.
This one change lifts time on page, lowers bounce, and gets you quoted more often. It costs one edit.
4. Link Your Own Pages Together
Internal linking is the most undervalued technique on this list because it is free, entirely within your control, and boring.
A navigation menu is not internal linking. A link inside a sentence, in the body copy, using words that describe where it goes, is what actually passes signal. Most business sites have almost none of these.
- Pick your most important page. The one that should rank for your main service.
- Find every other page that mentions that topic in its text.
- Add one link from each of those pages to the important one, inside a sentence, using the target phrase as the words you link.
- From the important page, link back down to the detail pages, again inside sentences.
Ten pages pointing at one page with consistent wording is a very clear signal about which page is the main one. You can do this in an afternoon and it costs nothing.
5. Make It Fast on a Phone, on Mobile Data
Not on your office wifi. On a phone, on mobile data, in the afternoon. That is how most of India will experience your site, and it is usually much worse than anyone building the site realises.
The fixes are nearly always the same four things, in this order:
- Compress the images. A 2 MB hero photo is the single most common cause of a slow page. Convert to WebP and most sites drop half their weight in an hour.
- Set width and height on every image. Without them the page jumps while loading, which counts against you and annoys people.
- Lazy load anything below the fold. One attribute per image.
- Cut unused scripts. Most sites carry three analytics tools, two chat widgets and a slider library they stopped using.
Speed will not lift a bad page into the top ten. It will absolutely stop a good page being held back, and it improves conversions whether or not rankings move.
6. Write a Title Somebody Wants to Click
Your title tag is doing two jobs. It tells Google what the page is about, and it is the advert that decides whether anyone picks you out of ten blue links.
Most businesses only do the first job. "Web Development Services | Company Name" is accurate and completely uninviting. Put the main phrase at the front, then add the one thing that makes yours worth clicking: a price, a timescale, a number.
Keep it under about sixty characters or Google trims it. Then go and look at your worst-performing page that ranks in the top twenty. A better title on a page already ranking at fifteen is the fastest win available in SEO, because the position is already earned.
7. Publish the Prices Everyone Else Hides
Almost every Indian service business writes "contact us for a quote". Which means that for every search containing the words cost, price, charges or fees, there is a gap in the results where an honest answer should be.
Filling that gap is unreasonably effective. You rank for a cluster of high-intent searches your competitors have abandoned, and you filter out the enquiries that were never going to afford you. The objection is always the same, that competitors will see the prices. They already know your prices. Your customers are the ones in the dark.
Take the monthly search volume. Assume you get about 25% of clicks at position one, 10% at three, 3% at seven, and roughly nothing below fifteen. Multiply by your enquiry-to-sale rate and your average sale value.
A phrase with 200 searches a month at position three brings about 20 visits. At a 10% enquiry rate and a 25% close, that is one sale every two months. If your average sale is ₹40,000 that keyword is worth chasing. If it is ₹800, it is not, no matter how easy it looks. Most people skip this and then wonder why a page that ranked changed nothing.
8. Earn a Few Real Links
Links still matter and there is no honest way to make this one easy. The good news is that you need far fewer than the industry implies.
For a local business competing on city-level searches, ten genuine links usually does it. Local news coverage, your trade association, a supplier or client who will list you, a genuine sponsorship, a directory people actually use. Slow, real, and they hold their value.
What does not work any more, and can actively hurt: bought link packages, guest posts on sites built only for guest posts, and directory blasts. If a link can be bought for ₹500 it is worth less than nothing.
| Technique | Effort | Time to show | Worth doing when |
|---|---|---|---|
| Match the intent | Ten minutes per page | Immediate on new pages | Always. Before writing anything. |
| One page, one job | A day to audit | Four to twelve weeks | You have more than about fifteen pages. |
| Answer in sixty words | One edit per page | Two to six weeks | Always. Cheapest fix here. |
| Internal links | An afternoon | Three to eight weeks | Always. Free and under your control. |
| Speed on mobile | A day of dev time | Slow, and indirect | Your pages take over four seconds on data. |
| Better titles | An hour | One to three weeks | You have pages ranking between five and twenty. |
| Publish prices | A week of nerve | Six to sixteen weeks | Competitors all say contact us for a quote. |
| Earn links | Ongoing, months | Three to nine months | Everything above is already done. |
Local SEO, Which Most Indian Businesses Need More Than Anything Above
If you serve one city, the map results sit above the normal ones and no amount of page work gets you into them. That is a separate system and it runs on three things.
- A complete Google Business Profile. Correct category, every service listed, real photos taken this year, hours that are actually right. Free, and most profiles are half empty.
- Reviews, arriving steadily. Twenty reviews spread over a year beats forty that all arrived in one week. Ask every satisfied customer, then ask once more.
- The same name, address and phone everywhere. If your address is written four different ways across four listings, that inconsistency is working against you.
| Signal | How much it matters | What to do this week |
|---|---|---|
| Primary category | Very high | Pick the most specific one that fits. Most profiles pick one too broad. |
| Review count and pace | Very high | Ask five customers. Steady beats sudden. |
| Distance from the searcher | High, and fixed | Nothing. This is why a shop three streets away outranks you. |
| Services listed on the profile | Medium to high | Add every service by name. Most profiles list two. |
| Photos, recent | Medium | Ten real photos taken this year. Not stock, not 2019. |
| Consistent name and address | Medium | Write it identically everywhere it appears online. |
| Replies to reviews | Low, but visible to buyers | Reply to all of them, especially the bad ones. |
For a local service business this is usually worth more in the first month than anything else on this page. Our own SEO work in Jaipur almost always starts here rather than with content, because the return arrives sooner and costs nothing.
What About AI Search
The practical answer is less dramatic than the headlines suggest: the things that get you quoted in an AI answer are mostly the things that were already good SEO.
Clear structure with real headings. The answer near the top. Specific figures and dates rather than vague claims. A named author with genuine experience. Content that says something a summary of ten other articles could not say.
The one real change in emphasis is originality. An article assembled from what already ranks has no reason to be cited, because the model already has that information from the sources you copied. First-hand numbers, your own pricing, your own results, a case you actually worked on, those are the parts nothing else can supply.
Techniques That Stopped Working
Time and money still go into these, so they are worth naming.
| Still being sold | Why it stopped working | Do this instead |
|---|---|---|
| Keyword density targets | Google stopped counting repetitions well over a decade ago, and it reads badly to humans. | Use the phrase once in the title, once in the opening, then write normally. |
| Meta keywords tag | Google confirmed it ignores the tag entirely, in 2009. It is still on thousands of Indian sites. | Nothing. Delete it. Spend the minute on the title instead. |
| Bulk directory submissions | Five hundred listings on sites no person visits. Worthless at best. | Five directories your customers actually use, filled in properly. |
| Word count targets | Padding a 700 word answer out to 2,500 makes it worse and buries the answer. | Write until the question is answered, then stop. |
| Exact-match anchors everywhere | Every link using identical wording looks engineered, because it is. | Vary it naturally. Link the words that genuinely describe the page. |
| Chasing a plugin score | A green 98 out of 100 measures agreement with that plugin. Google does not use it. | Check the search results for your phrase. That is the only scoreboard. |
| Monthly ranking reports | Positions vary by device, location and person. A single number is close to meaningless. | Track clicks and enquiries in Search Console, not positions. |
If an agency proposal still contains three of the rows in that left column, you have learned something useful about the proposal. It is also worth checking whether the basics are in place first, because a fast, well-built website makes every technique above work harder, and a slow one caps what any of them can achieve.
The most useful hour we spend on a new GIT Infosys client is almost never writing. It is listing every page with its target phrase in one column, because that is where the damage usually is.
On one recent audit, four pages carried the same phrase in their titles and their own canonical tags pointed at a URL that redirected back to the page. Nothing was going to rank until that was untangled, and no amount of new content would have fixed it. Check the plumbing before you buy furniture.
A Ninety Day Order of Work
Order matters more than effort here. Doing these in the wrong sequence wastes months.
- Weeks 1 to 2, audit. Every page, its target phrase, its title, its canonical. Find the duplicates and the broken plumbing. Fix nothing yet, just list it.
- Weeks 3 to 4, merge and redirect. Deal with pages competing against each other. This is subtraction and it is the highest value work in the quarter.
- Week 5, titles. Rewrite the titles on every page ranking between five and twenty. Fastest measurable win available.
- Week 6, the top sixty words. Put a direct answer at the top of every important page.
- Weeks 7 to 8, internal links. Wire your pages together properly, in body copy, with descriptive wording.
- Week 9, speed and local. Compress images, set dimensions, then complete the Business Profile and start asking for reviews.
- Weeks 10 to 12, write. Now write new pages, with intent checked before a word is drafted. Two good ones beat ten thin ones.
Notice that writing comes last. Most SEO plans start there, which is why so many of them produce a lot of content and very little movement. If you want help doing this properly on your own site, that is what our search engine optimisation work covers, in roughly this order.
The Short Version
- Check what Google already ranks for a phrase before writing the page.
- One page, one job. Merge the rest and redirect them.
- Answer the question in the first sixty words.
- Link your own pages together inside body copy.
- Test speed on a phone on mobile data, then compress the images.
- Rewrite titles on anything ranking between five and twenty.
- Publish your prices while your competitors hide theirs.
- Earn ten real links slowly and ignore anyone selling five hundred.
None of this is secret and most of it is free. The reason it works is not that it is clever. It is that it is dull enough that most of your competitors will keep buying reports instead.
Frequently Asked Questions
What are the best SEO techniques in 2026?
Matching search intent before you write, giving each page one target phrase, answering the question in the first sixty words, internal linking inside body copy, mobile speed, better title tags, publishing real prices, and earning a small number of genuine links.
Those eight cover most of what actually moves a page. Everything else is either upkeep or something an agency can put on a report.
How long does SEO take to show results?
Title changes on a page already ranking in the top twenty can move within one to three weeks. Fixing pages that compete with each other usually shows in four to twelve weeks.
Brand new pages take six to twelve months to earn anything meaningful, longer on a site with few links. Anyone promising page one in thirty days is either targeting phrases nobody searches or is not telling you the truth.
Is SEO still worth it now that AI answers are on top of search?
For commercial searches, yes. Nobody hires a contractor or buys a website from a summary box. Those searches still end on a page and that page still has to rank.
What lost value is thin informational content that a two line answer covers completely. If most of your traffic came from definitions and basic how-to posts, that traffic is genuinely going away.
What is the most common SEO mistake?
Two or three pages chasing the same phrase. Owners think it triples their chances. Google picks one, usually not the intended one, and all of them sit around position twenty.
The fix is subtraction. Pick the strongest page, merge the useful content into it, and 301 the others. It is the highest value work on most sites and it involves deleting things, which is why it rarely gets done.
How many backlinks do I need to rank?
Fewer than you have been told. For a local business competing on city-level searches, around ten genuine links is often enough, provided the pages themselves are sound.
Quality decides it. One link from a local news site or a trade body outweighs hundreds of directory listings, and bought link packages can now do active harm.
Should I publish my prices on my website?
Yes, in almost every case. Searches containing cost, price, charges and fees are high intent and most of your competitors have abandoned them by writing contact us for a quote.
It also filters your enquiries. You stop spending time on people who were never going to afford you, and the ones who call have already accepted the number.
Do I need to publish blog posts regularly to rank?
No. Publishing on a schedule for its own sake usually produces thin pages that compete with your own service pages.
Two genuinely useful pages a quarter, each aimed at a search you have checked, will beat twelve posts a month written to fill a calendar. Frequency was never the ranking factor people assumed.
Can I do SEO myself or should I hire someone?
The audit, the merging, the titles, the first sixty words and the internal links are all doable yourself and they are the highest value items on the list. None of them need technical skill, only patience and a spreadsheet.
Hire for the parts that need time you do not have: ongoing content, link work and technical fixes. Doing the free work first also means you can tell whether the person you hire knows what they are talking about.