Eywa has a 4.6★ rating from 742 guests and the best food photography in Banashankari. It also has ₹1.26 lakh of Google Ads spend aimed at an online checkout you never asked for. This is the 90-day plan to point it at the door instead.
You've been clear that Phase 1 has one objective: get more people through the door. No online sales, no delivery push, no e-commerce. Yet roughly ₹42,000 a month is being spent on Google Ads optimised toward an online checkout — a goal you never asked for, on a website that no longer exists.
This proposal is built around a single number: how many additional people physically walk into Eywa each month. Not impressions, reach, clicks, likes, followers or online orders. If a tactic here cannot be traced to a footstep on 13th Cross, it does not belong — and several things currently running fail that test badly.
It returns a Shopify "Store unavailable" error. That one dead URL is simultaneously the website link on the Google Business Profile, the landing page for every Google Ads campaign, and the product feed source for the Performance Max campaigns — one of which is already paused with the error "98% products are out of stock."
A full audit of the Google Ads account, Google Business Profile, Meta Ad Library, Instagram and the local competitive set — conducted 30 July 2026 on live data.
2 May – 30 July 2026 · 808,430 impressions · 41,900 clicks · "16,100 conversions" at ₹7.84 each.
16,100 conversions at ₹7.84 would be extraordinary for any cafe on earth. It isn't extraordinary — it's soft, auto-generated micro-actions. The deeper problem is what the account has been told to care about. Here is the funnel Google is optimising toward, for a business that does not sell online and never asked to:
Smart Bidding is being trained on a signal that structurally cannot occur — and that you would not want if it did. For three months it has been finding people who behave like online shoppers, not people who live near Banashankari and eat breakfast out.
| Campaign | Type | Budget | Impressions | Diagnosis |
|---|---|---|---|---|
| Eywa Cafe: Your Coffee Haven | Smart | ₹450/day | 439,805 | 54% of all account impressions on the weakest campaign type Google sells. No keyword, negative, audience or placement control. |
| Breakfast – Local store visits | PMax | ₹480/day | 175,413 | Optimisation score 78.3%. Feeding off a dead product feed. No occasion-specific asset groups. |
| Local store visits & promotions | PMax | — | — | Status: Eligible (Limited) — throttled, not serving to full potential. |
| Campaign #1 | PMax | ₹440/day | 89,809 | Unnamed default campaign. Limited by budget. Score 76.7%. No stated purpose. |
| Local store visits – PMax-3 | Paused | ₹150/day | 0 | Error: "98% products are out of stock." Dead Shopify feed. |
Total committed daily budget ₹1,670/day ≈ ₹50,100/month. Actual spend ran at roughly ₹42,000/month.
This is the clearest evidence the account was never built for your actual goal. Every signal measuring a real visit is partial or disabled; the online-behaviour goals run across the board.
For a walk-in cafe a phone call is among the highest-intent signals available: checking timings, booking a table, ordering a cake. Active in zero live campaigns.
"Shop visit" is the literal footfall metric Google offers. Active in half the account. The other half has no line of sight to the only outcome that matters.
WhatsApp and message enquiry — enormous in Indian F&B, especially for celebration cakes and grazing tables — is completely untracked.
Flagged by Google itself with an error state. An active goal in a broken configuration adds further noise to every bidding decision.
Online-behaviour goals remain active despite there being no online business to support them. These are the signals steering spend away from footfall, and they go first.
The Business Profile surfaces "Order online" alongside Swiggy and Zomato. Every delivery order is revenue at aggregator commission and a walk-in that didn't happen.
The highest review count of any cafe in the immediate area, and the owner actively responds — which Google rewards.
Every "Website" tap from Maps is a dead end. Free, immediate fix.
A genuine differentiator in local search that is currently doing no marketing work at all.
No Posts, no Offers, no Events. Google demonstrably favours actively maintained profiles in the local pack.
Two free surfaces sitting empty while competitors' unanswered questions rank instead.
Nothing from the profile is attributable, so its contribution can't be measured or defended.
The business appears as "Eywa – Bakehouse" on Maps, "Eywa Bakehouse" on Instagram, "Eywa" on Zomato and "Eywa Cafe" on Ads and Facebook. Reviews are split across Google (742), Justdial (822), EazyDiner (686), Zomato (118) and Swiggy. Google's entity resolution is being asked to guess. NAP consistency is a free, foundational fix.
Two active ads found in the Meta Ad Library, both launched 26–27 July 2026.
CTA: Call Now. Copy reads "loved by people from across Bengaluru… how far would you travel for a brownie?" That's destination-brand messaging. Phase 1 wants the person 3km away who hasn't tried it — not the person 15km away who won't drive on a Tuesday.
CTA: Contact Us, routing to fb.com. Good creative, wrong objective. A lead objective teaches Meta to find form-fillers, not people who walk into a cafe. No store-traffic objective, no radius targeting, no directions CTA anywhere.
The Meta Ad Library is saturated with a Dubai luxury real-estate development also called "Eywa" — R.Evolution's Water Canal tower, advertised heavily into India. Anyone bidding on "Eywa" without tight negatives and geo-fencing burns budget against property investors. A real, quantifiable leak, closed on day one.
Content quality is above market — proper food styling, lamination close-ups, staff features, workshop reels, well-structured highlights. Most cafes at this budget cannot produce creative this good. And essentially none of it is behind spend. That is the cheapest win available in this entire account.
| Cafe | Rating | Reviews | Price for one | Position |
|---|---|---|---|---|
| Third Wave Coffee | 4.6 | 616 | ₹200–400 | National chain, brand recall, price-led |
| Eywa – Bakehouse | 4.6 | 742 | ₹400–1,200 | Review leader, premium tier |
| Old Bangalore Cafe | 4.0 | 604 | ₹1–200 | Volume / value South Indian |
| Sukino Cafe & Kitchen | 4.4 | 281 | ₹200–1,000 | Closest direct overlap |
| Beankind Coffee | 4.8 | 198 | ₹200–400 | Highest rated, smallest base, drive-through |
Eywa leads on review volume but sits at 2–3× the price point of its nearest neighbours. The campaign cannot compete on the generic query "cafe near me" alone — that traffic compares ₹400 to ₹180 and bounces. Eywa has to be sold on occasion: weekend brunch, celebration cakes, work-friendly mornings, dessert after dinner, workshops. That is the entire keyword and creative thesis of this plan.
Phase 1 has exactly one KPI: get more people through the door. Every campaign, rupee and creative is judged against direction requests, calls, store visits and redeemed offers.
Someone in Banashankari is already searching "brunch near me" at 9:40am on Sunday. Eywa has no Search campaign to catch them. The fastest, highest-return fix in the account.
Most people within 3km still don't know Eywa exists. Meta's job is to put a croissant on their feed on a Friday evening until Eywa becomes the default answer to "where should we go?"
Paid acquisition is the expensive part. The Eywa Passport turns an ad-driven first visit into a habit — and generates free content while doing it.
Stated plainly, because you were specific about the goal. No online store. No e-commerce funnel. No delivery-aggregator promotion. No online-order objectives, add-to-cart events or purchase goals. All removed in Week 1 and none rebuilt. Delivery revenue is real revenue, but it is not footfall — and on aggregator commissions it can quietly cannibalise the walk-in you paid to create. If delivery or catering becomes a priority later, that's Phase 2 with its own budget.
Weighted toward Google, on the reasoning that Indian consumers overwhelmingly resolve "where do I eat" through Search and Maps rather than social discovery. Meta carries demand creation and retargeting.
Splits are directional and rebalanced weekly against actual cost-per-store-visit. Nothing is locked for 90 days.
The existing account isn't salvageable by tweaking. It gets restructured around one principle: every campaign must have a named job and a measurable footfall outcome.
The Smart campaign, "Campaign #1", and the dead-feed PMax-3. Together roughly 65% of impressions and near-zero controllable value.
Delete the Purchase, Add-to-basket, Begin-checkout and Page-view goals outright — not pause, delete. Fix the misconfigured Download goal. Activate Store Visits, Directions and Phone Call Leads across all campaigns.
A Merchant Center feed exists to sell things online. That isn't this business. Disconnected entirely, which also resolves the "out of stock" error permanently rather than patching it.
Block the Dubai "Eywa" real-estate collision, plus jobs, vacancy, recipe, franchise and wholesale terms. Geo-fence to South Bengaluru with presence targeting only — never "presence or interest".
Link the Business Profile to Ads, enable Location Extensions, Call Extensions with call tracking, and sitelinks to Menu, Cakes, Workshops and Directions.
GA4 and Google Tag configured correctly, UTM framework applied to every profile link, call tracking live from day one.
| Campaign | Type | Budget | Optimising for | Job to be done |
|---|---|---|---|---|
| EY | Search | Local Intent | Search | ₹14,000 | Store visits + Directions | Catch "cafe / brunch / breakfast near me" across a 5km radius. Exact and phrase match only. |
| EY | PMax | Local Store Visits | PMax | ₹10,000 | Store visits | Rebuilt as a pure Local campaign — no product feed, no online goals. Maps, YouTube and Discover pointed at directions and visits. |
| EY | Search | Cakes & Celebrations | Search | ₹5,000 | Calls → in-store collection | Highest-margin line, still a footfall play: cakes are ordered by phone and collected at the counter. Call-only ads on mobile. |
| EY | Search | Conquesting | Search | ₹4,000 | Directions | Bid on Third Wave, Beankind, Sukino and Blue Tokai within the local radius. Position on differentiation, not price. |
| EY | Search | Brand Defence | Search | ₹3,000 | Directions + Calls | Own "eywa", "eywa cafe", "eywa bakehouse" with airtight negatives against the Dubai property advertiser. |
Included in the retainer, and rupee for rupee the highest-return surface in this plan. It costs nothing to serve and sits directly between a hungry person and a decision.
What's-New, Offers and Events. Weekend brunch pushes, new bakes, workshop announcements. Google favours actively updated profiles in the local pack.
Every hero item — croissants, Turkish eggs, tiramisu, celebration cakes, workshops — loaded with photo, price and description. Free real estate almost nobody uses.
Pre-answer the fifteen questions that actually drive decisions: parking, wifi, vegan options, laptop-friendly, outdoor seating, cake lead time, group bookings.
QR code at billing plus a post-visit WhatsApp nudge. Target +40 new Google reviews a month while holding 4.6★. Review recency is a live ranking factor.
Geo-tagged, keyword-named, uploaded consistently rather than in bursts. Owner-photo freshness correlates with local pack visibility.
One name across Google, Justdial, Zomato, Swiggy, EazyDiner, Facebook and Instagram. Consolidates 2,500+ fragmented reviews under one entity.
Google catches people who already decided to eat out. Meta's job is to be the reason they decided. With ₹24,000 a month and creative this good, the constraint is objective selection — not budget.
Both current ads use lead-generation objectives — "Call Now" and "Contact Us" routing to fb.com. Everything moves to Store Traffic and Reach with a hard geographic radius. Meta's delivery algorithm optimises toward whatever you tell it to find. Right now it's being told to find form-fillers. We tell it to find neighbours.
| Campaign | Objective | Budget | Audience | Creative |
|---|---|---|---|---|
| EY | Store Traffic | 3km | Store Traffic | ₹8,000 | 3km pin radius around 13th Cross. 22–45. Presence-based. | Directions CTA, hero food, "five minutes from you" geo-hook |
| EY | Reels Reach | 5km Cold | Reach / Video Views | ₹6,000 | 5km. Specialty coffee, brunch, bakeries, food bloggers, Banashankari / Jayanagar / JP Nagar | Best organic reels, boosted only after clearing a retention threshold |
| EY | Retargeting | Warm | Store Traffic | ₹4,000 | IG engagers 365d, video viewers 25%+, profile visitors, customer list — capped to local radius | Offer-led, Passport signup, "you've seen us, come in". Directions CTA, not a link click |
| EY | Occasion Bursts | Store Traffic | ₹3,500 | Thu–Sun dayparting, weekend brunch and festival windows | Rotating: brunch, cakes, workshops, monsoon and festival creative |
| EY | Passport Push | Engagement | ₹2,500 | 3km radius plus all warm audiences | Manali trip reveal, leaderboard updates, winner stories |
POV breakfast content. Turkish eggs landing on wood, coffee poured, morning light. Targets the "where's brunch?" decision at the moment it's made.
Silent cold-open of dough being folded. No music for the first 1.5 seconds. Consistently the highest scroll-stop rate in food short-form.
Explicit geo-callout in the hook, naming streets and landmarks. Converts far better than generic "visit us" for radius campaigns.
Laptop-friendly, quiet, plug points, parking. Captures the weekday 11am–4pm dead zone — the single biggest margin opportunity in the day.
Celebration cakes and grazing tables. Highest-ticket product, currently almost invisible in paid. Drives calls and DMs, not just walk-ins.
Women-owned is a real differentiator. Face to camera, unpolished, phone-shot. Trust converts where food photography plateaus.
Baking and matcha workshops already exist in the highlights. Ticketed events that fill dead hours and generate weeks of UGC.
Real customers, real first bites. Native-feeling footage outperforms produced ad films on both cost and retention.
The Manali trip announcement, then a weekly leaderboard. Turns the loyalty programme into an ongoing content series.
Eywa already has a production partner. I don't want to replace them — I want to direct them. Beautiful footage that opens slowly is money set on fire.
Monthly content calendar, 12–15 scripted hooks, shot lists, retention-curve reviews on every published asset, and a data-led boosting decision so spend only goes behind creative that has already earned it organically. Fully optional; the core retainer stands alone without it.
That's the entire hook window. Visual and on-screen text hook must land simultaneously. No logo intros, ever — not one frame.
Motion resets attention. Text always on, since roughly 85% watch muted. 9:16 native only, safe zones respected.
Not a logo. The last frame should be an address and a map pin — the only thing you want remembered.
One shoot day should yield 12–15 usable vertical assets, not three. Production economics matter more than production values.
Over-produced footage reads as "ad" and gets scrolled. Phone-shot, handheld and imperfect consistently outperforms.
Publish organically, wait 48 hours, boost only what clears a retention benchmark. This alone typically cuts wasted creative spend by a third.
Contrarian opener. Names the competitor the viewer already knows, then earns the difference.
Geo-callout hook. Highest-converting format for radius-targeted store traffic.
Number and price hook. Directly answers the "isn't it expensive?" objection.
Occasion framing. Sells the moment, not the menu.
Curiosity gap plus craft. Justifies the premium price without ever defending it.
Pure sensory. Butter hitting the pan, espresso pulling, a knife through cheesecake.
Social proof as a hook, not a footnote.
Price-reveal format. Drives comments, which drives distribution.
Objection handled in the hook. Absurdly effective in Bengaluru.
Founder to camera, unpolished, phone-shot. Trust content converts where food content plateaus.
Ads buy the first visit. This buys the next twelve. A physical stamp-collection passport, handed over at billing, that turns an ad-driven stranger into a regular — and produces free content while it does.
Every guest receives a pocket-sized passport. Every visit earns a stamp. Milestones unlock rewards along the way — and the guest with the most stamps at the end of 90 days wins a two-night couple's trip to Manali.
1 stamp — any visit above ₹300
2 stamps — weekdays before 12pm (fills dead mornings)
2 stamps — weekdays 2–5pm (fills dead afternoons)
+1 — bring a friend who's never visited
+1 — order something new off the menu
+2 — attend a Sunday workshop
+1 — tagged Instagram story from the table
5 stamps — free specialty coffee
10 stamps — free dessert of choice
15 stamps — ₹500 voucher
20 stamps — Eywa merchandise, named on the wall
Most stamps in 90 days — Manali couple's trip, 2N/3D
Runners-up — ₹3,000 and ₹2,000 Eywa credit
Double stamps on weekday mornings and afternoons shift traffic into empty hours instead of discounting hours that were already full. Margin stays intact.
A weekly leaderboard on Stories, progress posts, milestone celebrations, a winner film. Roughly twelve content beats over 90 days at zero production cost.
Signup captures name, phone and birthday — becoming a WhatsApp list, a Meta Custom Audience and a Google Customer Match list. An owned asset that outlives the campaign.
The bring-a-friend bonus converts existing customers into an unpaid acquisition channel competing directly against paid CAC.
A public leaderboard makes visiting a game. The top twenty collectors will out-visit any loyalty card, and they do it publicly.
Every rupee of media now lands on a retention system instead of a one-off transaction. Lifetime value per acquired customer rises without raising spend.
Prize and print costs sit with Eywa, outside the ₹60,000 media budget — indicatively ₹40,000–50,000 for the Manali trip and ₹8,000–12,000 for passport printing across 90 days. I design the passport, build the mechanic, write the rules and run all promotion within the existing Meta budget. Folded into the marketing P&L it is comfortably cheaper per retained customer than buying the same visits with media.
Month 1 is repair and instrumentation. Month 2 is scale. Month 3 is compounding.
Not impressions. Not reach. Not likes. These are the only numbers that count in Phase 1.
| Metric | Today | Month 1 | Month 2 | Month 3 |
|---|---|---|---|---|
| Attributable store visits | Not tracked | Baseline set | 600–850 | 900–1,300 |
| Cost per store visit | Unknown | Baseline | ₹70–90 | ₹45–65 |
| GBP direction requests | Baseline | +30% | +65% | +100% |
| Calls from Google | Untracked | Tracking live | +50% | +90% |
| New Google reviews / month | ~15 | 25+ | 40+ | 50+ |
| Google rating | 4.6★ | ≥4.6★ | ≥4.6★ | ≥4.7★ |
| Passport signups | — | 150+ | 400+ | 700+ |
| Instagram followers | 3,831 | 4,300 | 5,200 | 6,500+ |
Return modelled at roughly ₹650 average spend per head against a combined ₹80,000 monthly outlay. Projections are directional, based on the audited account data and local benchmarks — not a guarantee. Store-visit measurement requires Google's modelled threshold to be met, expected within 30–45 days of correct implementation.
A short WhatsApp summary every Monday — spend, store visits, cost per visit, what changed and why.
Full dashboard review with a 45-minute call. Channel breakdown, creative performance, GBP insights, next month's plan.
A live Looker Studio dashboard you can open at any hour. No black box, no waiting for a deck.
Three things sit outside my control. Two are quick. None of them block launch.
Because the goal is footfall and not online sales, the dead site matters far less than it first appears. It is not a lost sales channel — it's missing ad infrastructure. It costs two things only: Search landing quality, which nudges CPCs up, and remarketing audience collection. It does not block launch and it does not cap footfall. Traffic routes to the Business Profile, phone and WhatsApp — which is where a walk-in decision actually gets made.
A separate agency is already restoring eywacafe.com. Good — let them. If it slips, a simple one-page site would tighten Search performance, but it is genuinely optional here and I'd rather say so than sell you something Phase 1 doesn't need.
Admin on Google Ads, Google Business Profile, Meta Business Manager and Instagram. GBP admin in particular — I could not reach the dashboard during this audit and it is central to the plan.
The Passport lives or dies at the billing counter. Staff need a twenty-minute briefing and a simple script. I'll write it and run the session; the team needs to actually use it.
Transparent and unbundled, so Eywa can start with the core and add the module only if it earns its place.
Direction only — production stays with Eywa's existing partner.
Media spend is billed directly by Google and Meta to Eywa's own payment methods — I never hold, invoice or mark up ad budget. Service fee payable monthly in advance. GST applicable as per prevailing rates. The optional module is quoted separately and is not included above.
90-day initial sprint. Local search and review signals take six to eight weeks to compound — anything shorter measures noise. Rolling monthly thereafter, 30 days' notice either side.
Eywa owns every account, asset, audience and dataset created during this engagement. No lock-in, no hostage accounts.
Once footfall compounds, the same infrastructure extends to delivery, corporate catering, workshop ticketing or a second outlet. Recommended on what the data says at day 90.
Eywa has already built the hard part — a genuinely good cafe that 742 people rated 4.6 stars. Everything in this document is plumbing. Give me 90 days and a measurement stack pointed at one goal, and the same ₹60,000 will do work it has never done before.
Approve & begin Week 1Week 1 can begin within 48 hours of approval.
Proposal prepared 30 July 2026 · Audit conducted on live Google Ads, Google Business Profile, Meta Ad Library and Instagram data · Valid 30 days