Digital Marketing Skills: What Indian Job Ads Actually Ask For

AI made the left column cheap. That is exactly why the right column decides who gets hired.

Most skill lists for this field are written by people selling a course. This one is read off the job ads: seven current Indian listings, agency and in-house, counted for what they actually demand. If you are still working out what the job is, start with our guide for beginners, which owns that ground. This page covers the skill stack, the order, and the proof.

What seven Indian job listings actually asked for

Read as a set, the seven agree on a narrower stack than any syllabus suggests. Analytics is universal, paid platforms close behind, and AI is almost absent from the requirements whatever the discourse says.

• Performance Marketing Executive, ad agency, Andheri, Mumbai (Pink Sky HR Services, 2 to 3 years, ₹25,000 to ₹35,000 a month advertised): Google Ads, Meta Ads Manager, YouTube and DSPs, plus "Google Analytics, Tag Manager, Data Studio, SEMrush" and "proficiency in Excel, dashboards, and reporting". Listing on Indeed India.
• Performance Marketing Specialist, Andheri, Mumbai (EveryMedia Technologies, 3 to 4 years, ₹40,000 to ₹50,000 advertised): the same platforms, plus Tag Manager and stakeholder management. Listing.
• Jr. Performance Marketing Executive, Mumbai (Seventy9 Media, 0 to 1.5 years, ₹20,000 to ₹25,000 advertised, expired when we read it): setup on both ad platforms, "strong Excel/Google Sheets skills", certification "optional but preferred". Listing.
• Sr. SEO Executive and SEO Executive, Andheri East, Mumbai (Realatte, ₹30,000 to ₹40,000 and ₹20,000 to ₹40,000 advertised): both ask for "Google Analytics, Google Search Console, SEMrush, Ahrefs, Moz" and written communication, and neither mentions AI, AEO or GEO. Senior listing · junior listing.
• Digital Marketing Executive, in-house, Kandivali East, Mumbai (Palram India, 2 years, ₹35,000 to ₹40,000 advertised): wider and shallower. WordPress, on-page and off-page SEO, HubSpot or Zoho, "strong communication, copywriting", Canva, with analytics and paid ads only desirable. Listing.
• Digital Marketing Executive, Bengaluru (Oneskai, 2 to 3 years, ₹18,000 to ₹25,000 advertised): the only one asking for "familiarity with AEO and GEO concepts", "AI productivity tools such as ChatGPT, Google Gemini" and "prompt engineering for marketing use cases". Listing.

At the entry level it shifts. On Internshala's Mumbai jobs page, read the same day, "Social Media Marketing" is the most-repeated tag and Google Analytics appears on almost none of the junior postings. Entry-level ads hire for posting. Two years in, they hire for measurement.

Two caveats. Seven listings is what we could verify in one sitting, not a survey, and the counts are our own tally. Every rupee figure is the band advertised on that listing.

Table stakes, differentiators, and how long each honestly takes

Table stakes are the skills whose absence ends the conversation: one ad platform, GA4, a spreadsheet, and a sentence a client can read. Differentiators move you from shortlisted to hired: conversion tracking you set up yourself, a keyword map with intent attached, and the judgement to call a number noise.

The channels themselves are covered in our guide to the types of digital marketing, and one job role end to end in our social media manager guide. The table below is the other axis: the skill, not the channel or the title.

Where AI has changed the job, and where it has not

AI has changed the mechanics of campaign construction and the speed of first drafts. It has not changed what you are accountable for, and it has created no new technical requirement on the SEO side.

On paid, the change is structural. Google's documentation for AI Max in Search campaigns describes keywordless matching, automatic text customisation from your landing pages, and final URL expansion, while the advertiser keeps brand controls, URL exclusions and a search terms report with an AI Max match type (Google Ads Help, accessed 24 September 2026). Writing 200 exact-match keywords is gone. Reading what the machine matched to, and excluding it, is the job.

On search and content, Google's line has not moved. Its spam policies name "using generative AI tools or other similar tools to generate many pages without adding value for users" as scaled content abuse (Search Essentials, last updated 28 August 2026).

So if a course sells you an "AI SEO" module built on special files or markup, check the source. Google states there are "no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary" and that "you don't need to create new machine readable files, AI text files, or markup" (AI features and your website, last updated 10 December 2025).

The analytical skills most juniors skip

The gap is not tools. Most candidates can open GA4. The gap is deciding what a number means before quoting it.

• Defining the conversion, not inheriting it. If GA4 counts a form view as a lead, every number downstream is wrong. Our GA4 setup guide covers the event configuration.
• Reading the search terms report, not the keyword list. Keywords are what you asked for. Search terms are what you bought.
• Knowing when a change is too small to read. Forty clicks and one extra lead is not a 100% improvement. It is one lead.
• Segmenting before concluding. Mobile and desktop, new and returning, branded and non-branded. A flat average hides most of what happened.

The writing skill nobody lists but everyone needs

Six of the seven listings asked for "excellent communication". None said writing, and writing is what they meant. Three kinds, in order of frequency.

1. The status email on a bad week. What happened, what you changed, what you will know by Friday. Short, unhedged, no adjectives.
2. The one-line reason inside a report. Not "CPL increased 22%", but "CPL increased because we turned off the branded campaign on the 12th".
3. Ad copy and subject lines, the compressed kind where one word decides the click.

A model drafts all three faster than you can. It cannot know that the founder hates the word "solutions", or that last month's dip was a warehouse outage. Supplying that is the skill. The craft side sits in our content writing guide.

How to build a portfolio with no clients

You do not need clients. You need artefacts: things you made, with a decision and an outcome attached. Take Aarti, a hypothetical final-year student in Mumbai with nothing but coursework. Five artefacts, all buildable in a month, none needing anyone's permission.

• A live site with GA4 and Tag Manager on it. Any topic. It proves you configured measurement, not that you can blog.
• One paid campaign you funded. ₹2,000 to ₹3,000 of your own money on one Meta or Google campaign beats any simulation, because a rejected ad and a wasted ₹400 are both real.
• An audit of a local business that did not ask for one. Two pages: what is broken on their Google Business Profile, the fix you would make first, what you expect it to move. A work sample, not a cold pitch.
• A keyword map with intent labels. Forty keywords, each pointed at a page that exists or should.
• Four weekly reports in a row. Same format, one commentary line each. This is the artefact that gets the second interview, because it is the actual job.

Then make it findable: our LinkedIn profile guide covers the surface those artefacts sit on. Most people in this field learned inside a team rather than alone, which is the work our services do daily for brands.

Certifications worth the time, and the ones that are not

Only two of the seven listings mentioned certification, both as optional or nice-to-have. Treat a certificate as vocabulary, not evidence: it teaches you what the buttons are called so an interviewer does not have to.

• Google Ads certifications (Skillshop): worth it. Free, 80% to pass, 75 minutes, valid one year, in Search, Display, Video, Shopping and Apps (Google Ads Help). Take Search, and take it early, since its value is the vocabulary.
• HubSpot Academy's content marketing certification: worth it for the structure. Free, around eight and a half hours across 14 lessons (HubSpot Academy). Nobody in India hires on it, but it is a decent syllabus.
• Meta's professional certifications: only when someone else is paying. Media Buying Professional is a proctored exam maintained annually through recertification (Credly badge page). A paid, expiring badge is a poor use of a fresher's own money.
• Long paid "advanced digital marketing" programmes: ask one question first. Does the fee include a live ad account with real budget you run yourself, and a mentor who reviews your decisions? If not, you are buying the free Skillshop syllabus at a markup.

A realistic six-month learning order

Depth first, breadth later: one channel done properly beats six skimmed, because only depth survives a follow-up question. This assumes roughly ten hours a week.

1. Month 1: measurement before marketing. Put up one small site. Install GA4 and Tag Manager yourself, define two conversion events, verify them in preview mode, connect Search Console. Break it once and fix it.
2. Month 2: one paid channel, real money. Meta or Google, not both. Fund one campaign. Run it four weeks. Write down every change you made, the date, and what you expected.
3. Month 3: search and keywords. Build the 40-keyword map with intent labels, publish three pages against it, and watch Search Console for impressions, which arrive well before clicks.
4. Month 4: the second paid channel, and the certification. Add whichever platform you skipped, then sit the free Skillshop Search exam, once you have an account to attach the words to.
5. Month 5: reporting and writing. Four weekly reports in one tab format, twenty ad variants for one product, and one page rewritten with a note on what changed.
6. Month 6: package and apply. One page, five artefacts, each with its decision and result. Apply to twenty roles. Expect the agency interview to go deeper on platforms and the in-house one wider on tools.

For 2027, our expectation is that AEO and AI-assistant visibility moves from the one-in-seven column towards the middle of the list, and that reading automated campaign output becomes more of the paid job than building campaigns is. We put no number on either.

What separates the juniors who get hired (our view)

Opinion, formed from hiring juniors at an Indian agency. Not data, and we have no study to point at.

The candidates who get offers can describe one decision they made and why, without reaching for a framework. "I turned off the interest targeting on day nine because CPL had doubled and audience overlap was 40%." That sentence does more than any certificate, because it proves someone watched something and acted.

The ones who do not get offers bring a list of tools and a folder of certificates, and when asked what happened when they used them, describe the tool again. The second-best answer in an interview is "I do not know". The worst is a rehearsed definition.

One more, and it costs nothing: bring a number you were wrong about. Saying so out loud is the fastest trust you will build with a hiring manager, and it separates the two piles more reliably than anything on a CV.