What is the difference between WhatsApp Native Flows and a custom chatbot in India 2026?

WhatsApp Native Flows are Meta-hosted forms and UI screens that render directly inside the WhatsApp client — defined by a flow_json document, optionally posting submissions to a Laravel endpoint_uri you control. A custom chatbot is a server-side conversation graph (nodes, edges, per-contact state) that drives the conversation message-by-message through the Cloud API. Native Flows are the right answer for single-screen-or-short structured data capture (lead form, appointment booking, KYC field collection); custom chatbots are the right answer for multi-step branching logic, conditional routing, third-party API enrichment, payment collection and long-running automations.

Are Meta Native Flows cheaper than running a custom chatbot in India?

On Meta conversation pricing, they are billed identically — a Flow message falls under the same utility, marketing or authentication category as any other message, and the Meta India 1 January 2026 rate revision applies equally. The cost difference shows up in build-time and infrastructure. Native Flows have lower build cost for simple form capture (no backend state engine needed) but cap out fast: anything that needs conditional branching beyond two screens, payment retries, third-party CRM enrichment or async waiting periods needs a custom chatbot. RichAutomate ships both in the box at INR 0 platform fee plus INR 0.10 per conversation markup, so the build-vs-buy tradeoff is the only one that matters.

Can I use Native Flows for DPDP Act 2023 consent collection?

Yes, and this is one of the strongest single use cases. A Native Flow with a checkbox-and-text-confirm screen renders consent capture inside WhatsApp itself, posts the audit-trail payload to your endpoint_uri, and keeps the consent record co-located with the contact. Under the November 2024 draft DPDP Rules, in-app consent capture with audit trail is materially stronger evidence than a one-line opt-in checkbox on a web form. For complex consent ladders (purpose-by-purpose toggles, withdrawal mechanics, child-subject parental consent), a custom chatbot gives you the branching and conditional logic that Flows cannot.

When should an Indian D2C brand pick Native Flows over a custom chatbot?

Pick Native Flows when the entire interaction fits in 1 to 3 screens of structured input with no conditional routing — examples include a returns request form, an NPS survey, a basic appointment booking with a fixed slot picker, or a single-purpose lead form for a paid ad campaign. Pick a custom chatbot when the conversation needs to branch on user input, call a third-party API mid-flow (inventory check, UPI mandate creation, KYC verification), wait for an async event (payment confirmation, agent response, SLA timer), or carry per-contact state across multiple days. Most production D2C journeys use both: a custom chatbot orchestrates the conversation, and Flows are invoked as embedded forms inside specific nodes.

Do Native Flows work with the Meta India 1 January 2026 rate hike the same way?

Yes. As of June 2026, a Flow message is priced under the same conversation category as the wrapping message — utility, marketing or authentication. The rate revision does not introduce a Flow-specific surcharge. The strategic implication is that businesses cannot use Native Flows as a way to dodge the marketing-category rate increase — if your inbound is marketing-tier, the Flow inherits that price. See our full analysis at the January 2026 rate-hike impact piece.

How do RichAutomate and AiSensy / WATI / Interakt differ on Native Flow support?

As of June 2026, RichAutomate, AiSensy, WATI, Interakt and Gupshup all support Meta Native Flows on the Cloud API. The differentiation is on (a) whether the platform exposes a visual flow_json editor or you must paste raw JSON, (b) whether the endpoint_uri submission lands in your own database with audit trail or stays in the vendor silo, and (c) whether you can mix Native Flows with a custom chatbot in the same journey. RichAutomate ships both engines natively and a visual builder for both; most competitors expose one or the other cleanly, not both.

WhatsApp Native Flows vs Custom Chatbot: When to Use Which (India 2026)

Most Indian WhatsApp teams in 2026 deploy the wrong flow engine for the job and discover it only after the third production iteration. They pick a custom chatbot for a 30-second lead form (over-engineered, slow to ship), or they pick Native Flows for a multi-step KYC-plus-payment journey (caps out at screen three, ends up in a brittle workaround). The choice is not religious — it is a six-criterion decision matrix where each criterion is independently testable. This piece walks through the matrix, the cost-per-conversation math under the Meta India 1 January 2026 rate revision, the DPDP Act 2023 implications under the November 2024 draft Rules, and the three vertical patterns where the answer flips from one engine to the other.

Direct answer (June 2026). Use Meta Native Flows when the entire interaction fits in 1 to 3 screens of structured input with no conditional routing — lead forms, returns requests, NPS surveys, appointment booking with fixed slots. Use a custom chatbot when the conversation needs branching logic, third-party API calls mid-flow, async waiting, payment retries, or per-contact state across days. Most production journeys in India use both: the custom chatbot orchestrates the conversation and invokes Native Flows as embedded forms inside specific nodes. RichAutomate ships both engines on a single canvas at INR 0 platform fee plus INR 0.10 per conversation.

Native Flows and custom chatbots solve different problems

WhatsApp Native Flows (also called Meta Flows) are a Meta-hosted UI primitive: you define a flow_json document describing screens, fields, validation and a terminal action, and Meta renders that as a native form inside the WhatsApp client. Submission payloads are POSTed to a Laravel endpoint_uri you control. The reference implementation lives in Meta's WhatsApp Flows documentation.

A custom chatbot is the opposite shape. The conversation lives as a graph of nodes (send message, wait for reply, branch on condition, call API, schedule delayed action) with per-contact state persisted on your server. RichAutomate's FlowExecutionService drives one message at a time through the Meta Cloud API v24.0, evaluates triggers on inbound webhooks, and uses delayed-job queues for time-based steps. The two systems do not compete — they compose. A Native Flow can be invoked from inside a custom-chatbot node when the journey hits a structured-data-capture step.

For broader background read our pillar on the best WhatsApp Business API India 2026 and the in-depth piece on WhatsApp Business API setup in India step-by-step.

The 6-criterion decision matrix

Run your use case through these six criteria, in order. The first one that gives an unambiguous answer wins.

Interaction depth: single screen vs multi-step

Native Flows win if the entire interaction is 1 to 3 screens of structured input with a single submit. Examples: returns request, NPS rating, lead form for a paid ad campaign, appointment booking with a fixed slot picker. Custom chatbot wins if the interaction needs more than three turns of back-and-forth, dynamic message content based on previous answers, or any non-form messaging (rich media, sequential confirmations, conditional follow-ups).

Conditional branching requirement

Native Flows support basic field-level validation and a small set of dynamic data substitutions, but they do not support conversational branching ("if user picks A, go to screen X; if B, go to screen Y; if C, escalate to a human agent"). The moment your decision tree has more than two paths, you need a custom chatbot. This is the criterion that flips most often once a project moves from MVP to production.

Third-party API enrichment mid-flow

Native Flows can fetch dynamic data via the endpoint_uri at screen-render time (a stock list, an available-slot list, an account balance), and Meta refreshed these dynamic-data hooks through 2025. But anything that needs to chain multiple API calls, retry on failure, or branch on the API response, belongs in a custom chatbot. RBI-regulated lenders pulling CIBIL data mid-flow, IRDAI-regulated insurers running underwriting checks, and SEBI-registered research analysts validating KYC against the depository — all custom chatbot territory.

Async waiting: payment, SLA timer, agent reply

A Native Flow is a synchronous interaction — the user fills, submits, gets a response. The moment your journey needs to wait for a Razorpay UPI mandate confirmation, a 30-minute SLA timer, an agent reply on a shared inbox, or a downstream system to acknowledge an order, you need a custom chatbot with a delayed-job queue. RichAutomate's flow engine uses Redis-backed delayed jobs for exactly this shape; see visual flow builder feature.

DPDP consent audit-trail control

Both engines can capture DPDP Act 2023 consent. Native Flows are stronger for one-screen simple consent capture — the submission lands in your endpoint with a structured payload that includes the contact, the consent purpose, the timestamp and the WhatsApp message ID, which is materially stronger evidence than a web-form checkbox under the November 2024 draft DPDP Rules. Custom chatbots win for granular purpose-by-purpose toggles, withdrawal mechanics and conditional consent ladders. See DPDP consent manager checklist.

Build-time vs operate-time trade-off

Native Flows have lower build-time for the matching use case (single-screen form): no backend state engine, no node graph, no delayed-job infrastructure. Custom chatbots have higher build-time but lower operate-time once you cross the second or third iteration — adding a new branch is cheaper than re-defining a flow_json. As a heuristic, if you expect to iterate the journey weekly for the next six months, the chatbot is cheaper in total. If it is a fire-and-forget form for a single campaign, the Native Flow is cheaper.

Cost economics under the Meta India 1 January 2026 rate hike

One sentence answer. A Native Flow message and a custom-chatbot message are billed identically by Meta — both inherit the conversation category (utility, marketing or authentication) of the wrapping conversation, and the Meta India 1 January 2026 conversation-rate revision applies equally to both.

The cost difference is at the platform and infrastructure layer, not the Meta layer. Below is the per-conversation cost model as of June 2026 for the three most common India BSPs plus RichAutomate.

Platform	Native Flows support	Custom chatbot support	Platform fee (per conv)	Build-time tooling
RichAutomate	Visual flow_json editor + endpoint_uri	Visual graph builder + delayed jobs	INR 0.10 flat markup	Both engines on one canvas
AiSensy	Yes (flow_json paste)	Limited graph (linear)	Included in fixed tier	Two separate editors
WATI	Yes (visual)	Limited (rule-based)	Included in USD tier	Native Flows-first
Interakt	Yes (visual)	Yes (Haptik engine)	Included in fixed tier	Two separate editors
Gupshup	Yes (enterprise)	Yes (BotScript)	Volume-committed	Code-heavy, enterprise

For per-vendor head-to-heads see RichAutomate vs AiSensy, RichAutomate vs WATI, RichAutomate vs Interakt, RichAutomate vs Gupshup and the broader pricing analysis at is WATI or AiSensy cheaper for Indian SMB 2026. The Meta-side cost framework is in Meta India January 2026 rate-hike impact.

Three vertical patterns where the answer flips

Use cases that look identical on the surface often need different engines once you examine the actual workflow.

D2C: lead capture vs abandoned cart

Lead capture from a Meta paid ad — Native Flow. One screen, four fields, submit to CRM. Abandoned cart recovery — custom chatbot. The journey needs to wait 2 hours after cart abandonment, send a reminder, branch on whether the user replies (offer discount vs offer help), call Shopify for stock check, push to Razorpay for a payment link with a 24-hour expiry, and retry on payment failure.

BFSI: balance enquiry vs loan application

Balance enquiry — Native Flow with dynamic data hook to the core banking API, single screen, two fields (account number + OTP), submit returns the balance. Loan application — custom chatbot. KYC verification, CIBIL pull, eligibility decisioning, document collection across multiple turns, regulatory disclosures per RBI norms, NACH mandate setup, async approval. See India regulation pillar.

Healthcare: appointment booking vs symptom triage

Appointment booking with a fixed slot picker — Native Flow. Dynamic-data hook for slot availability, single screen, submit creates the booking. Symptom triage — custom chatbot. Conditional branching across 15-plus symptom dimensions, escalation to a human clinician on red-flag responses, async wait for prescription approval, integration with the EHR. The DPDP-Act-2023 health-data provisions push the consent ladder into chatbot territory for anything beyond a single-screen consent.

The hybrid pattern most production journeys use

In practice, most India production WhatsApp journeys shipped through 2025 and into 2026 are hybrids. The custom chatbot is the spine of the conversation; it handles routing, state, async waiting, third-party calls and DPDP consent ladder. At specific moments — KYC field collection, NPS survey, appointment slot selection, address capture — the chatbot invokes a Native Flow as the data-entry primitive. The Flow returns its submission to the chatbot via the endpoint_uri, which advances the chatbot graph to the next node. This pattern uses Native Flows for what they are best at (in-WhatsApp structured forms with native UI) and chatbots for what they are best at (orchestration, branching, persistence).

For the related architectural discussion on Cloud API vs on-premise, see Cloud API vs on-premise WhatsApp Business India 2026. For RBI / IRDAI / TRAI references in regulated journeys, see IRDAI and TRAI source pages.

Why 2024 guides got this wrong

Most pre-2026 content treats Native Flows as a feature checkbox on a BSP pricing page rather than as a distinct architectural primitive. That framing made sense in 2023 when Native Flows were newly released and most BSPs had not yet exposed visual editors. As of June 2026 the situation is reversed: every credible India BSP supports Native Flows on the Cloud API, and the differentiation is now (a) whether the platform exposes a visual flow_json editor, (b) whether submissions land in your own database with audit trail or in the vendor silo, and (c) whether you can mix Native Flows with a custom chatbot in the same journey. The strategic question for buyers in 2026 is not "does this vendor support Native Flows" — it is "does this vendor support the hybrid pattern." See the broader competitive context in our State of WhatsApp BSP India Q2 2026 research.

What to do next

Bring your top-three WhatsApp use cases (lead form, recovery journey, support flow) to a short call. We will run each through the six-criterion matrix live, map them to the right engine (Native Flow, custom chatbot, or hybrid) and model the per-conversation cost under the Meta India January 2026 rate card on your real volume. Book a 30-minute fit call, or message us on WhatsApp at +91 74349 01027.