craftitstudios

Digital operations studio

Anyone can launch.Staying liveis the work.

/ site, listings, search, outreach, one dashboard /

Most studios hand you a website and wish you luck. We take the part that comes after — the listings, the search presence, the outbound, the numbers — and run it as one system you can actually see into.

Start a conversationSee what we run

99.9%

Uptime we commit to, with service credits

1 hr

First response when production is down

4–6 wks

From audit to live

Live

Everything in one view

Traffic, revenue, listings and outbound in a single dashboard, not five tabs and a spreadsheet.

Look inside Studio OS

The gap nobody quotes for

A prototype answers “can this exist”.
Production answers “can this take money at 3am”.

You can absolutely get a working site out of an afternoon and an AI. That part genuinely got easy. What did not get easy is everything between a thing that runs on your laptop and a thing your business can stand on.

    When traffic arrives

    Weekend buildWorks fine for the four people who tried it

    Production

    Load tested, cached, and alerting before users notice

    The technical version
    Budgets set on Core Web Vitals and enforced in CI, so a regression fails the build. CDN caching with a documented invalidation path. Synthetic checks running against checkout and signup every few minutes, paging a human on failure rather than filling a dashboard nobody opens.

    When money moves

    Weekend buildPayment succeeds, and that is where it ends

    Production

    Payments reconcile, refunds work, failures are recoverable

    The technical version
    Idempotency keys on every payment path so a double-submit does not double-charge. Webhook handlers that verify signatures and survive retries. Refund and chargeback flows built before launch, not after the first dispute. Ledger reconciled against the gateway so the number in your dashboard matches the number in your bank.

    When search looks

    Weekend buildIndexed eventually, ranked for nothing

    Production

    Crawlable, structured, and aimed at real queries

    The technical version
    Server-rendered content rather than client-side hydration crawlers have to be trusted to execute. Canonical discipline, no orphaned templates, sitemap generated from a route manifest. JSON-LD for Organization, Product, FAQ and Breadcrumb. Log-file review to confirm what crawlers actually fetch versus what we assume shipped.

    When you email people

    Weekend buildSent from a fresh domain, straight to spam

    Production

    Authenticated, warmed, and landing in the inbox

    The technical version
    SPF, DKIM and DMARC aligned on a dedicated sending subdomain, warmed over weeks. Suppression lists honoured across every channel. One-click unsubscribe with List-Unsubscribe headers. Get this wrong once and the domain reputation is difficult to recover — which is why it happens before the first send, not after.

    When a customer asks about data

    Weekend buildA privacy policy copied from another site

    Production

    Documents that survive a real security questionnaire

    The technical version
    Processing agreement, named sub-processors, defined retention, and a route for data subject requests with a stated response window. Consent gating that blocks non-essential scripts before consent rather than loading them and asking afterwards. The version most sites run would not survive being read carefully.

    When it breaks at 3am

    Weekend buildYou find out from a customer, on Monday

    Production

    Someone is already on it

    The technical version
    Error tracking with source maps so a stack trace points at real code. Structured logs that can be queried rather than scrolled. A rollback that is one command and gets rehearsed. Defined escalation with response windows written into the SLA instead of implied.

None of this is exotic. It is just the work that gets skipped, because it is invisible right up until the moment it is the only thing that matters.

What we run

Five things, turning together

Each of these is sold separately by someone. Bought separately, the seams between them are where the work leaks out.

  1. The site

    A storefront that loads fast on a cheap phone and stays up when traffic actually arrives.

    • Design system, not a template
    • Staging environment
    • Rollback in one command
    • Uptime and error alerting
    The technical version
    Next.js on managed infrastructure with preview deploys per change. Core Web Vitals budgeted in CI — a regression fails the build rather than getting noticed a month later. Structured logging, error tracking with source maps, and synthetic checks against the paths that take money. Images served in modern formats at the sizes actually requested, fonts self-hosted and subset.

  2. Product listings

    Your catalogue entered properly once, then kept correct everywhere it appears.

    • Product data model
    • Bulk import and validation
    • Variant and stock logic
    • Channel sync
    The technical version
    A normalised product schema with required fields enforced at the boundary, so incomplete listings cannot reach production. Image pipeline handles derivatives and alt text. Product schema.org markup emitted per listing for rich results. Feed generation for marketplace and shopping channels, diffed on each run so a silent field change does not quietly delist you.

  3. Search presence

    Pages built so search engines can read them, and content aimed at what people actually type.

    • Technical audit and fixes
    • Keyword and intent map
    • Content calendar
    • Monthly position reporting
    The technical version
    Crawlability first: clean internal linking, canonical discipline, no orphaned templates, sitemap generated from a route manifest rather than maintained by hand. JSON-LD for Organization, Product, FAQ and Breadcrumb. Server-rendered content — not client-side hydration search engines have to be trusted to execute. Log-file review to see what crawlers actually fetch versus what we think we published.

  4. Reaching people

    Calls, messages and email that land in front of real prospects instead of in a spam folder.

    • Target list, sourced and scrubbed
    • Sequenced calls and messages
    • Newsletter with real deliverability
    • Reply handling
    The technical version
    Domain authentication done properly — SPF, DKIM and DMARC aligned, a dedicated sending subdomain warmed over weeks rather than blasted on day one. Suppression lists honoured across every channel, opt-outs processed immediately. Do-not-call scrubbing before dialling. Double opt-in on the newsletter with one-click unsubscribe and List-Unsubscribe headers, because the alternative is burning the sending domain permanently.

  5. Studio OS

    One dashboard showing what is happening, what it earned, and what needs a decision.

    • Traffic and revenue in one view
    • Funnel and drop-off
    • Integrations already wired
    • Weekly digest
    The technical version
    Event pipeline with a defined schema, so metrics stay comparable across months instead of drifting each time something is instrumented. Analytics, payments, email and support tools connected on our side — you are not handed API keys and a tutorial. Data export available on request in a format you can actually use, because your numbers should never be held hostage by a dashboard.

Studio OS

The part that makes us not an agency

Agencies send reports. We give you the dashboard the reports would have been made from — live, and the same one we work out of.

What happened, and what it earned

Traffic, revenue, conversion and spend on one screen, over the same date range, without exporting anything.

Events run through a defined schema, so a metric means the same thing in March as it did in January. Most dashboards drift because instrumentation is added ad hoc; this one does not, because the schema is versioned.

Connected on our side. You are not handed a set of API keys and a documentation link.

  • Analytics

    Where visitors come from and what they do

  • Payments

    Revenue, refunds and failed charges

  • Email & newsletter

    Sends, opens, unsubscribes, deliverability

  • Search console

    Queries, positions and crawl errors

  • Support inbox

    Conversations tied to the customer record

  • Error tracking

    What broke, for whom, with a stack trace

  • Product feeds

    Catalogue pushed to marketplaces

  • Call tooling

    Dials, connects and outcomes

How it goes

Four phases, and you always know which one you are in

  1. Audit

    Week 1

    We go through what exists — site, listings, search position, sending reputation, tooling — and write down what is actually wrong. You get that document whether or not you continue.

    • Technical audit
    • Search and listings baseline
    • Priority order with reasoning

  2. Build

    Weeks 2–6

    The site and the data model get built properly, on staging, where you can see it. Nothing goes live because a deadline arrived.

    • Design system
    • Production site on staging
    • Catalogue imported and validated
    • Analytics instrumented

  3. Launch

    Week 6–7

    Cutover with a rehearsed rollback, monitoring live before traffic arrives, and the compliance surface in place rather than promised.

    • Cutover plan
    • Monitoring and alerting
    • Legal pages live
    • Sending domain warmed

  4. Operate

    Ongoing

    The part that is the actual job. Content ships, outbound runs, numbers get reviewed, and you get a weekly digest short enough that you will read it.

    • Weekly digest
    • Monthly review call
    • Content and outbound cadence
    • Studio OS access

Who this fits

Two ways in

Which describes you

You want the thing to work without learning why it does.

You will never be asked to pick a hosting provider or debug a DNS record. We make the technical decisions, write down what we decided in language you can read, and you approve direction rather than implementation.

  • One point of contact, not a ticket queue
  • Plain-language weekly digest
  • Decisions explained, not delegated back to you
  • Nothing that requires you to log into a server, ever

Who this is

Two people, both of whom you will actually talk to

No account manager between you and the work. That is the advantage and it is also the limitation — we take on a small number of operations at a time because there are two of us.

Straight to both of us: founders@craftitstudios.com

Evidence

What we can show you before you commit

We are early, so instead of a wall of logos you may not recognise, here is what we will put in front of you first.

  • A real audit of your current setup

    Not a lead-generation PDF. The actual list of what is broken, ordered by what it costs you, written before any money changes hands.

  • The technical standard, in writing

    Performance budgets, uptime target, response windows, security posture. Specific enough to hold us to.

  • A walkthrough of Studio OS

    The live dashboard, with real data in it, so you can judge whether it would actually tell you anything useful.

  • The code we have written

    Both our GitHub profiles are linked below. Read the actual work rather than taking a claim about it — that is a harder form of evidence than a testimonial and we would rather be judged on it.

We will not show you client logos we do not have, or growth charts from someone else's case study. When there are results worth showing, they will appear here with a name attached and permission to check.

Engagement

Priced as an operation, not as a project

A build has an end date. An operation does not, which is why this is a monthly retainer with a one-off build phase rather than a lump sum and a handover.

Audit

Fixed fee

one-off, quoted up front

Find out what is actually wrong before deciding anything.

  • Technical and performance audit
  • Search and listings baseline
  • Deliverability and sending review
  • Prioritised findings document
  • Ninety-minute walkthrough call

Credited in full against the build if you continue.

Ask for a quote

Operation

Most take this

Monthly retainer

quoted after the audit

The whole thing, run continuously.

  • Everything in the five pillars
  • Studio OS access for your team
  • Weekly digest, monthly review call
  • Content and outbound cadence
  • Monitoring, alerting and on-call
  • Compliance surface maintained

Build phase quoted separately. No long lock-in — 30 days' notice, no exit fee.

Talk to sales

Embedded

Monthly retainer

capacity permitting

For teams that want us inside their workflow.

  • Everything in Operation
  • Shared repository and code review
  • Your tooling and your process
  • Tighter response windows
  • Quarterly technical planning

There are two of us, so this has a hard ceiling of one or two clients at a time.

Ask about availability

Why there is no price on this page

Because it would be a guess, and a guess wide enough to be useless. The work depends on what already exists, how large the catalogue is, and whether your sending domain has reputation damage to repair — that last one alone can be weeks. The audit is a fixed fee quoted before you commit to anything, so the only thing you agree to up front is finding out.

All figures quoted in INR and exclusive of applicable taxes. Billing cycle and payment terms are agreed in writing in your proposal.

Questions

The ones worth asking

I already built something with AI. Is that wasted?

No, and it is genuinely useful. A working prototype tells us what you actually want, which is normally the slowest part to establish. We keep the decisions and rebuild the parts that will not survive real traffic — usually the data model, the payment paths and anything touching search.

Why not just hire an agency and a marketer?

You can, and for some businesses that is the right answer. The difference is the seam: an agency owns the site, a marketer owns the traffic, and the gap between them is where things quietly fail. We take both sides plus the dashboard that shows whether either is working.

Do I own what you build?

Yes, on final payment for the build phase — code, design files, content and data. That goes in your written proposal rather than being left to goodwill. If we stop working together you get an export and the repository, not a hostage negotiation.

What happens if I want to leave?

You give 30 days' notice, we hand over the repository, the data export, the runbooks and the credentials, and we run a transition call with whoever takes over. No exit fee.

There are only two of you. Is that a risk?

Yes, and it is the honest objection to working with us. You get the people doing the work with no account manager in between, and we cap how many operations we run at once so that stays true. But we are two people, not a department — if you need twenty-four-hour coverage across time zones or a team that survives one of us being ill for a fortnight, say so early and we will tell you whether we can commit to it.

Can you guarantee search rankings?

No, and anyone who does is either lying or defining success narrowly enough to be meaningless. We can commit to the technical work, the content cadence and honest position reporting. Where you land depends on your market, which we do not control.

Is the cold outreach going to get my domain blacklisted?

Not if it is done properly, which is the entire reason we do authentication, warming and suppression before the first send. We scrub against do-not-call registers, honour opt-outs immediately, and pause sending automatically when complaint rates approach the thresholds providers act on.

How technical do I need to be?

Not at all. There is a version of every update written in plain language, and a technical version underneath it if you or someone on your team wants the specifics. Nobody is expected to read a deployment log.

What do you not do?

Brand strategy from scratch, paid ad management, video production, and anything requiring a physical presence. We will tell you when something is outside what we do rather than learning it on your budget.

How about a conversation

Tell us what you aretrying to run.

No pitch deck. A conversation about what exists now, what it needs to do, and whether we are the right people for it.

We reply within one working day. If we are not the right fit we will say so in that reply rather than booking a call to say it slowly.

You

Optional

What exists now, and what it needs to do.

Where you think you are