How to choose a digital marketing agency (and avoid the wrong one)
Most businesses choose an agency badly, and the reason is understandable: you are being asked to evaluate technical work you cannot yet judge. The proposals all look similar, the promises all sound confident, and price becomes the deciding factor by default. This guide gives you the questions that actually separate a capable agency from a convincing one.
In this guide
The twelve questions that matter
About the work
- What exactly will you deliver each month? A real answer has quantities — keywords tracked, articles published, campaigns managed, calls scheduled. "We will improve your SEO" is not a deliverable.
- Who actually does the work? Many agencies resell to freelancers. That is not automatically bad, but you should know who touches your account and whether they change.
- Can I see work you did for a business like mine? Sector experience matters more than a large client list. Property search behaves nothing like e-commerce.
- What does month three look like? If they cannot describe realistic milestones, they have not planned past the sale.
About results
- What will you report, and how often? The report should show enquiries and cost per lead, not just impressions and rankings.
- Which metrics do you consider a failure? An agency willing to define failure is thinking about accountability. Most avoid this question entirely.
- Have you ever lost a client, and why? The answer tells you a great deal about honesty. Everyone has lost clients.
- What have you tried that did not work? Anyone who has done this seriously has a list.
About the relationship
- Who is my point of contact, and can I speak to them now? Sales handover to an unnamed junior is the most common complaint in this industry.
- What do you need from me each month? Good agencies are clear about what they need — approvals, content, photos. Vague answers mean delays later blamed on you.
- What is your notice period? Three months minimum for SEO is reasonable. Twelve-month lock-ins with no exit are not.
- What happens to my accounts if we part ways? Covered in detail below, and it is the question most worth asking.
Red flags worth walking away from
- Guaranteed rankings. Nobody controls Google's results. A guarantee means either meaningless keywords or risky tactics.
- Results promised in 30 days. Paid ads can move fast; SEO cannot. Compressed timelines signal either inexperience or dishonesty.
- No written scope. If deliverables are not documented, disputes are inevitable and you will lose them.
- They want to own your ad account. Your Google Ads and Meta accounts should be yours, with them granted access. Not the reverse.
- Reporting only on vanity metrics. Impressions and follower counts without enquiry data usually means enquiry data is not flattering.
- Pressure to sign immediately. Limited-time discounts on a service relationship are a sales tactic, not a business reality.
- They cannot explain their own work simply. Jargon is often used to avoid scrutiny. Anyone who understands SEO can explain it plainly.
Who owns what when you leave
This is the question businesses ask too late, usually while trying to leave. Settle it before signing.
| Asset | Should belong to | Why it matters |
|---|---|---|
| Google Ads account | You | Contains all historical performance data used for optimisation |
| Meta Business Manager | You | Losing it means rebuilding audiences and pixel history |
| Website files and hosting | You | Otherwise your site is hostage |
| Domain name | You | Registered in your name, always |
| Google Analytics & Search Console | You | Years of data, impossible to recreate |
| Google Business Profile | You | Primary ownership must be yours |
| Content produced | You | Should be work-for-hire, transferring on payment |
| Social media accounts | You | Created under your business email, not theirs |
A reasonable agency will have no problem with any of this — they work inside your accounts with granted access. Resistance on any line above tells you what the relationship will look like when it ends.
Contract terms to negotiate
- Minimum term. Three months for SEO is fair and genuinely necessary. Six is negotiable. Twelve with no exit clause is not.
- Notice period. Thirty days after the minimum term is standard.
- Scope changes. How are additional requests priced? Get the hourly or per-item rate upfront.
- Reporting cadence. Monthly report plus a call. Put it in writing or it quietly becomes quarterly.
- Ad spend handling. Billed directly to your account by Google and Meta. State it explicitly.
- GST clarity. Whether quoted figures include 18% GST. Ambiguity here causes the first argument.
- Pause clause. Can you pause for a month during a slow season? Useful for seasonal businesses.
How to run a fair trial
If you are uncertain, structure a small paid engagement rather than committing to a year. A paid SEO audit is the best test available — it costs little, produces something concrete, and shows you exactly how they think.
What a good audit looks like
- Specific problems identified, with evidence and screenshots
- Prioritised by impact, not listed alphabetically from a tool
- Written so you can understand it without a glossary
- Includes things that are working, not only failures
- Ends with a clear plan, not a sales pitch
An audit that is obviously exported from a tool with a logo added tells you exactly what the retainer will look like. One that shows someone actually examined your site and understood your business tells you the same thing, in the opposite direction.
The best signal is not what an agency promises. It is whether they tell you something you did not want to hear during the sales conversation.
References & further reading
Official sources and tools referenced in this guide:
Frequently asked questions
What should I ask a digital marketing agency before hiring?
Focus on four areas: exactly what is delivered monthly with quantities, who does the work, what happens to your accounts if you leave, and what they consider failure. An agency that answers all four directly is usually worth talking to further.
Should I hire a local agency or a remote one?
For local SEO and Google Business Profile work, an agency that understands your city genuinely helps — they know how the market behaves. For SEO, ads and content, location matters far less than sector experience and communication quality.
Is a cheaper agency always worse?
Not automatically, but price constrains scope. At ₹5,000 a month there is not enough time for technical work, content and link building all three. If a quote is much lower than others, ask precisely which parts are excluded — something always is.
How long should I commit initially?
Three months is fair for SEO because results genuinely take that long. Ads can be evaluated in six to eight weeks. Avoid twelve-month contracts with no exit clause unless the pricing is significantly better and you have checked references.
What if the agency does not deliver results?
Define "results" in writing before starting — agreed metrics, agreed timeline. Without that, disagreements become subjective. A monthly report showing enquiries and cost per lead, reviewed on a call, catches problems in month two rather than month ten.
Want this done for your business?
Send us your website and we will come back with a prioritised list of what to fix first — no obligation, no sales pressure.
Get a free digital audit
Tell us your goals and current setup. We will send back a specific, prioritised list of what to fix first — no obligation.
Request Free Audit →Or call +91 79825 40118 · +91 74510 05142
How an engagement runs
Six stages, in this order. We do not skip ahead because a client is impatient — the sequence exists because each stage depends on the one before it.
Discovery and audit
Everything currently running is reviewed together — search, paid, social, website, follow-up. Channels are almost never failing independently; the gaps between them are usually where the money leaks.
One strategy document
A single plan covering every active channel with a stated role for each. Without it, channels compete for the same budget while claiming credit for the same enquiries.
Shared research layer
Keyword and audience research done once and used by every channel. Paying two agencies to research the same business twice is the most common hidden cost of a split setup.
Sequenced rollout
Paid launches first because it produces data fastest. Organic and content build underneath it. Social runs alongside to make both cheaper. Turning everything on at once makes attribution impossible.
Conversion infrastructure
Landing pages, forms, tracking and CRM routing. Sending more traffic to a site that does not convert is an expensive way to discover the site does not convert.
One report, one call
Consolidated monthly reporting that leads with enquiries and cost per enquiry, with channel detail underneath for anyone who wants it.
The mistakes we see most often
Every one of these comes from an account or a site we have actually inherited. None of them are hypothetical.
Splitting channels across vendors
Each optimises their slice, nobody owns the gaps, and research gets paid for twice. The gaps between channels are where enquiries leak.
Turning everything on at once
Simultaneous launch across five channels makes attribution impossible. Sequenced rollout tells you what actually worked.
Judging channels in isolation
Social rarely closes the sale directly; it makes search and paid cheaper. Measured alone it looks like failure, measured in context it often subsidises everything else.
No conversion infrastructure
More traffic to a site that does not convert is an expensive way to confirm the site does not convert. Fix the destination before increasing volume.
Changing strategy every month
Compounding channels need at least two quarters to show their shape. Agencies overhauling the plan every reporting cycle are usually managing anxiety, not performance.
Questions worth asking
Ask these of us and of everyone else you are considering anywhere. The answers separate agencies faster than any proposal does.
If an answer is vague, that is the answer. Vagueness at the sales stage becomes vagueness in reporting.
- — Who owns the accounts, domain and data if we leave?
- — What exactly is delivered each month, in writing?
- — Do you mark up ad spend or media cost?
- — Who actually does the work, and where do they sit?
- — What is the notice period, and what happens on exit?
- — Can I speak to a client you stopped working with?
- — What would make you tell me to stop spending?
What results realistically look like
Honest timelines, based on engagements we have actually run rather than what sounds persuasive in a pitch.
Foundation
Audit, fixes and setup. Little visible movement. This is the stage clients find hardest, and skipping it is why most engagements underperform later.
First signals
Early movement on long-tail terms, campaigns stabilising, first attributable enquiries. Enough data to know what is working and what is not.
Compounding
Commercial keywords move, cost per enquiry falls as the account matures, and content published earlier starts ranking. This is where the work pays back.
Scale or stop
By now the numbers are clear. Either we scale what works, or we tell you honestly that the channel is not right for your business.
Terms you will hear, explained
Agencies hide behind jargon. Here is what the words actually mean, so you can hold any vendor to account — including us.
Attribution
Assigning credit for an enquiry across the channels that touched it. Last-click flatters paid and undervalues everything upstream.
CAC
Customer acquisition cost — total spend divided by customers won. The number that decides whether marketing is working.
Full funnel
Covering awareness, consideration and decision rather than only the final click.
Marketing stack
The combined tools running your marketing. Fragmented stacks lose data at every handoff.
Incrementality
Whether a channel produced sales that would not have happened anyway. Rarely measured, frequently assumed.
Things clients ask before signing
Can we start with one channel and expand?
That is usually what we recommend. Two channels, then expand once there is data showing where the next rupee performs best. Buying everything on day one is rarely the efficient move.
How is this cheaper than separate vendors?
The research and reporting layer is shared instead of repeated. Combined engagements typically land 20–30% below the sum of the same services bought individually.
Who is our point of contact?
One account strategist who owns the whole engagement. You do not brief a different person per channel.
What contract length?
Monthly rolling after an initial three-month period, which exists because compounding channels need a fair run to show anything. No annual lock-in.
What if it is not working?
We tell you, with numbers, and either change approach or recommend you stop. An agency that never delivers bad news is not reporting honestly.