Start with our conflict of interest
Last Agency sells SEO, organic social and performance marketing. Our bundles are ₹99,000, ₹75,000 and ₹1,75,000 a month depending on which channels you take. Every argument for consolidation on this page puts money in our pocket, and every argument for splitting takes it out.
So treat the pro-consolidation half sceptically and the pro-splitting half as something we'd rather not have written. The honest answer is that this is a measurement question dressed up as a procurement question, and the right call changes with your budget and with how weird one of your channels is.
What one agency actually buys you
Not efficiency. The savings from consolidation are real but small — one kickoff, one reporting cadence, one set of access requests. What you actually buy is the ability to hold a single number and have one person answer for it.
That matters more than it sounds. Most marketing arguments in Indian companies aren't about strategy; they're about whose report is right. One supplier collapses that argument into one document.
- Search data crosses channels for free. The queries converting in Google Ads should be steering next quarter's content plan, and the non-brand pages already ranking should be excluded from paid bidding. Two agencies can do this. They rarely do, because it means one of them shrinking their own scope.
- One creative brief. Whatever line is winning on paid social should be the headline on the landing page and the H1 the SEO team writes against. Split suppliers produce three different value propositions and call it channel nuance.
- One person to blame. Underrated. When an agency knows it cannot deflect to the other vendor, the quality of the monthly conversation changes immediately.
| What has to work | One agency | A specialist per channel |
|---|---|---|
| The baseline | One frozen number covering every channel. Movement is unambiguous. | Each supplier freezes a different number, usually the one that flatters it. |
| Attribution | One model, one property, one definition of a lead. | Three dashboards, three answers, and nobody wrong enough to be fired. |
| A month where nothing moved | One explanation, from the one person accountable. | Two suppliers pointing at each other while the quarter closes. |
| Depth in a hard channel | Capped at the generalist's weakest discipline. | As deep as the market has, if you shop properly. |
| Marking its own homework | A genuine risk — the supplier picks the model and reports the score. | Largely solved. Each one's numbers get checked by the others. |
| Cost | One retainer. Usually cheaper per channel. | Three minimum fees stacked on top of one another. |
Where splitting genuinely wins
These aren't hypotheticals. Each one is a case where we would tell you to go elsewhere for that channel, and where a generalist claiming competence is guessing.
- A channel that is more than half your budget. If ₹8 lakh of a ₹12 lakh monthly spend sits in Google and Meta ads, that channel deserves a supplier whose whole business is that channel. The tail can't wag the dog.
- Marketplace advertising. Amazon and Flipkart ad platforms have their own auction logic, their own reporting and their own seasonal rhythm. Search competence does not carry over.
- App store optimisation. Different store, different index, different ranking inputs. It shares a name with SEO and almost nothing else.
- Regulated categories. Pharmaceutical, financial services, education-abroad and healthcare advertising carry compliance obligations that a generalist will discover after the ad is disapproved or, worse, after it isn't.
- A migration or a re-platform in flight. Technical SEO on a large JavaScript build is a specialist engagement with a defined end date, not a line item inside a marketing retainer.
- Regional-language creative at scale. If you need Tamil, Telugu, Marathi and Bengali production every month, hire the people who write in those languages rather than the agency that briefs a vendor who does.
The double count: one lead, two invoices, two agencies claiming it
This is the failure nobody warns you about at the pitch, and it turns up in almost every split-supplier account we're asked to review.
A buyer sees a paid social ad on Monday, searches your brand on Wednesday, lands on a blog post from organic on Friday, and fills the form. Your paid agency reports the lead, because the ad platform attributes conversions back to the ad interaction that preceded them. Your SEO agency reports the same lead, because the last non-direct click before the form was organic. Both reports are internally correct. Together they describe two leads that do not exist.
The mechanics are worth knowing rather than trusting. Google Ads attributes a conversion to the ad click that preceded it, and its conversion counting options decide whether one ad interaction can produce one conversion or several. Analytics, meanwhile, distributes credit across the path according to whichever attribution model the property is set to. Two systems, two philosophies, one form fill.
At small volumes this is an annoyance. At 200 leads a month across three suppliers it is a number in a board deck that nobody in the room can reproduce.
The setup we insist on before we take all three channels
Consolidation's failure mode is the mirror image: one supplier chooses the attribution model, builds the dashboard and reports the score. That's a supplier marking its own homework, and no amount of goodwill fixes a structural conflict.
So we make the following non-negotiable, and you should make it non-negotiable with any agency running more than one channel.
- You own the accounts. Google Analytics, Google Tag Manager, Search Console, Google Ads, Meta Business Manager — all created under your company's ownership, with us added as users. Not the other way round.
- One Analytics property, one definition of a key event. Not a separate property per channel and not a second GTM container that a previous agency still controls.
- One attribution model, written into the contract. Pick it, name it, and don't change it mid-engagement. Switching models re-scores history, which is the oldest trick in agency reporting.
- Every campaign URL tagged consistently. UTM parameters agreed as a naming convention up front, using Google's own URL builder conventions, so the channel labels in your reports mean one thing all year.
- A self-reported source field on the enquiry form. "How did you hear about us?", free text, mandatory. It disagrees with analytics constantly, and the disagreement is the useful part.
- The CRM is the system of record. Reports reconcile to it monthly. If a channel report and the CRM disagree, the CRM wins and the discrepancy gets explained in writing.
- Brand and non-brand split on every organic number. Otherwise a paid campaign that lifts brand search quietly pays the SEO team's bonus — the same conflict, one supplier smaller. There's more on holding rankings or traffic to account if you want the reporting clause.
The budget below which the question doesn't matter
Every agency has a floor beneath which it staffs your account with whoever is free. Ours is ₹40,000 a month for a small site, ₹75,000 for real SEO. A competent paid team starts around ₹40,000 a month for management; organic social starts at ₹30,000.
Stack three minimums and you're at roughly ₹1,10,000 a month before anyone has done anything ambitious — and each supplier is getting the least interesting third of your budget. At that level you are not buying three specialists. You are buying three account executives who each have eleven other clients.
Our working rule: below about ₹2,00,000 a month of total retainer, consolidate. The coordination tax on splitting eats more than the specialism gains. Above ₹4,00,000 a month, splitting starts to pay for itself, because each supplier's slice is now large enough to interest their senior people. Between the two, it depends on how weird your hardest channel is.
That's a rule of thumb from what we see in the Indian mid-market, not a law. If your paid spend is ₹50 lakh a month against a ₹1 lakh retainer, ignore the rule and hire the paid specialist first — the cost comparison of running SEO against ads is a different question from who runs each one.
The verdict
Consolidate if your channels serve one buyer, your total retainer is under roughly ₹2,00,000 a month, and no single channel is strange enough to need a dedicated practice. Take the shared baseline and the single accountable person, and pay for it by owning every account and every tag yourself.
Split if one channel is more than half your budget, or if it's marketplace ads, app store work, or a regulated category where getting it wrong is a legal problem rather than a performance one. Then accept the cost: you now run the reconciliation, and it is your job, not theirs.
The setup that fails either way is the middle one — three suppliers, no shared baseline, each with its own tracking, and a founder who assembles the monthly picture out of three PDFs on a Sunday night. If that's you, the number of agencies isn't the problem. The missing baseline is. Set one first, then decide how many suppliers it takes to move it.