MotionReach Data Services

MotionReach operates 500 in-taxi screens today and will scale to 10,000 over the next 24 months. Those screens are currently sold as an advertising channel with a single metric: the number of times an ad played. Proof-of-play is a delivery receipt, not a business outcome, and it caps what a client will pay per screen-hour.

Data Services turns the same hardware into a continuous data-collection network. Every play already happens somewhere, at some time, in front of someone, on a journey between two points. Capturing that context, with passenger consent and a fair value exchange, converts an unmeasured impression into a measured moment, and a measured moment into an attributable outcome.

The vertical has two halves. The capture module builds the pipeline: consented passenger profiles, GPS-at-play, journey context, touch and QR engagement, and client-supplied conversion feeds. The commercialisation module turns that pipeline into four distinct revenue lines: measurement attached to media, syndicated benchmarks, bespoke insight studies, and audience licensing for programmatic buying.

This is directly aligned with the investor narrative: offline-to-online attribution and programmatic are both flagged as 'this round' capabilities, and 35% of the US$3M is allocated to MotionVision engineering. Data Services is the concrete product plan behind that line item.

The network today reports plays. A campaign that delivered 4.54M verified impressions is a strong delivery story, but it answers none of the three questions clients ask after the invoice: who saw it, did they act, and was it worth it.

Three structural assets make this solvable now rather than later. We own the hardware, so we control the sensor and the screen. We own marketing execution, so we sit close enough to the client to negotiate conversion data. And we are small enough at 500 screens to instrument the network properly before the 10,000-screen rollout locks in a worse design.

MotionReach Data Services exists to make in-car mobile media the most measurable offline channel in Southeast Asia, and to make MotionReach the owner of the metric that proves it.

Positioning statement

For brands and agencies who spend offline and cannot prove it worked, MotionReach Data Services is the measurement layer of the mobility network, the only source that connects a specific ad play, in a specific vehicle, on a specific journey, to a named audience profile and a verified downstream action.

Three positioning pillars

• Consent-first. Passengers choose what they share and are paid for it in value, not tricked out of it. This is the moat: consented first-party data cannot be scraped or resold by a competitor.
• Journey-native. Our data is not a panel survey about mobility; it is mobility. Origin, destination, route, time and dwell are captured as a by-product of the medium itself.
• Outcome-bound. Every metric we publish must be traceable to a client business outcome, or it does not go in the product.

The currency: MotionReach Conversion Index (MRCI)

A single normalised 0–100 score per campaign, comparable across category, creative and geography, built from four weighted components. It is the artefact that turns Data Services from a reporting feature into an industry standard.

The capture module is a five-layer data model. Each layer is independently valuable and independently shippable, and later layers depend on earlier ones being clean.

4.1 Data taxonomy

4.2 Passenger profile & the consent product

The passenger profile is a product in its own right, not a form. It runs on the touchscreen during the natural dead time of a journey and is designed around a simple bargain: more disclosure, more reward, always reversible.

Consent UX principles

• Plain-language Thai and English, one screen, no dark patterns, no pre-ticked boxes.
• Granular toggles by purpose: network optimisation, brand insight, personalised offers, loyalty linking.
• One-tap withdrawal at any time in the ride, plus a QR to a self-serve privacy portal after the ride.
• Data minimisation by default: we store bands and zones, not exact ages and exact addresses.
• Every consent event is logged immutably with version of the notice shown. That is the audit trail.

4.3 Incentive economics

The reward is a real cost line and must be engineered, not improvised. The goal is to have brands, not MotionReach, fund most of it: a voucher from an advertiser is inventory to them and cash-equivalent to the passenger.

Target unit economics: blended cost per completed T2 profile under THB 12, against an insight and measurement revenue per profiled ride that should exceed THB 40 by H2. Cost per profile is a board-level KPI from month one.

4.4 Survey and insight capture

• Maximum three questions per ride, single-tap answers, under 20 seconds, designed for a Bangkok traffic-light gap, not a research lab.
• A rotating question bank: a fixed core tracker (category usage, brand funnel, mood) plus client-sponsored modules.
• Quality controls: speeding detection, straight-lining detection, duplicate-device suppression, attention-check items, and per-screen response caps.
• Weighting to a Bangkok population frame using census and mobility benchmarks so results are projectable, not just descriptive.
• Field capacity model: at 500 screens and ~14 rides/screen/day, a 12% T2 rate yields roughly 25,000 profiled rides per month, enough for national topline cuts within one month and district cuts within a quarter.

4.5 QR and offline-to-online attribution

1. 01Every play renders a QR encoding a unique play token: screen ID, creative ID, timestamp, geo cell and session hash.
2. 02Scanning hits the MotionReach redirect service, which records the scan and forwards to the client destination with tagged parameters.
3. 03Online outcomes come back through a client pixel, server-side conversion API, or a shared campaign parameter, matched on the play token.
4. 04In-store outcomes come back through POS transaction files, loyalty redemption of a MotionReach-issued code, or geo-visit measurement against store polygons.
5. 05Every campaign reserves a holdout: matched screens or matched districts that never carry the creative, providing the counterfactual.

4.6 Client-side data partnerships

Loyalty is a distinct wedge for banks, fuel, pharmacy, grocery and airline categories: a passenger enters their loyalty number on the screen, is served member-only offers for the rest of the journey, and the brand gets a clean, consented, in-journey activation channel with closed-loop redemption data.

4.7 PDPA and governance

• Lawful basis: explicit consent for profile and personalisation; legitimate interest only for aggregated network operations.
• Consent register with notice versioning, timestamp, and tier, retrievable per data subject.
• Retention: raw personal records 12 months, aggregated and de-identified analytics indefinitely; automated purge jobs, not manual cleanups.
• Aggregation floor: no cut published below the sample thresholds; small cells suppressed rather than rounded.
• Data subject rights portal: access, correction, deletion, withdrawal, served within statutory timelines.
• Contracts: DPA with every advertiser receiving data, data-sharing agreement with fleet operators, driver acknowledgement covering vehicle telemetry.
• A named Data Protection Officer and an annual third-party privacy audit once T3 linking goes live.

Most offline measurement stops at correlation: spend went up, sales went up, therefore the ads worked. Clients have learned to discount that. The framework below moves MotionReach up the evidence ladder step by step, so that every claim we make is backed by the strongest method the available data supports.

5.1 The metric chain

5.2 Methods by data maturity

5.3 Design rules that protect causality

• Holdouts are the default, not an upsell. Every campaign above a spend threshold reserves 10–15% of matched inventory as control.
• Randomise at the unit we can control, screen or district, and match on historical delivery, daypart mix and passenger profile.
• Pre-register the primary metric and the analysis window before the campaign runs. No post-hoc metric shopping.
• Always report a confidence interval alongside the point estimate; a lift without an interval is marketing, not measurement.
• Separate observed facts from modelled estimates visually and in writing, in every deliverable.

Data Services should not be given away to protect media revenue. It is a higher-margin business than media, it defends the media business, and it should be priced and staffed as a separate P&L from day one.

6.1 Measurement attach

The entry product. Every media proposal includes a measurement pack; a minority of clients opt out. It converts a media transaction into an annual relationship because the client cannot compare this year's campaign to last year's without staying on the platform.

6.2 Syndicated insight, Mobility Pulse

• Free topline: a quarterly public report on Bangkok mobility, ad engagement norms and category trends. This is the marketing engine and the credibility builder.
• Paid cuts sold as subscriptions: by category (auto, beauty, banking, F&B, retail), by ad format, by audience profile, by district and daypart.
• Benchmark tables are the hook. A brand cannot know if a 2.1% scan rate is good without the category norm, and we own the norm.
• Publication cadence and an annual 'State of Mobile Media' flagship give the sales team a reason to be in front of every CMO once a year.

6.3 Bespoke insight studies

Pre-campaign: audience sizing, journey mapping, creative testing on-network. Post-campaign: full attribution readout, creative diagnostics, next-flight recommendations. This is the service that makes MotionReach a strategic partner rather than a vendor, and it is the single best predictor of campaign renewal.

6.4 Audience and licensing

Consented profiles enable programmatic delivery: the passenger's declared interests select the ad loop for their journey. Advertisers pay a CPM premium for targeted, verified, in-journey audiences. Longer term, aggregated mobility and category-demand data has licensing value to retail, property, financial and public-sector buyers who need to understand movement in Thai cities.

6.5 Commercial targets

H1 · Months 0–6: Instrument the network

• Ship the consent and profile module on the touchscreen (T0–T2), in Thai and English.
• Log a GPS fix, heading and speed with every play; define trip origin/destination zones.
• Deploy unique per-play QR tokens and the redirect and scan-capture service.
• Stand up the data warehouse, the event schema, and the internal network dashboard v1.
• Run three paid attribution pilots with existing clients who can supply conversion data.
• Publish the PDPA framework, consent register and privacy portal.

Exit criteria: 95% of plays carry a valid geo fix; T1 opt-in above 20%; three pilots delivered with a signed-off lift number; zero privacy incidents.

H2 · Months 6–18: Reach critical mass and productise

• Add T3 loyalty linking and the rewards ledger.
• Build the weighting engine and the Bangkok population frame; validate projectability.
• Launch the client-facing self-serve dashboard with MRCI scoring.
• Launch Mobility Pulse syndicated reports and the subscription product.
• Ship programmatic audience segments into the ad server.
• Automate holdout design and lift reporting so measurement scales without analyst hours.

Exit criteria: 100k+ profiled rides per month; publishable district-level cuts; 12 syndicated subscribers; measurement attached to over half of media revenue.

H3 · Months 18–24+: Own the currency

• Independent methodological validation and published back-testing of MRCI.
• Agency adoption programme: get MRCI written into media plans and post-campaign templates.
• Data licensing product and API for approved partners.
• Replicate the capture spec into new SEA markets as the fleet expands.

Dependencies and sequencing risks

The fastest way to destroy a data business is to publish a confident number from thirty responses. These thresholds are a hard product rule enforced in the dashboard, not a guideline.

Quality controls

• Response validity scoring on every survey completion; invalid responses excluded before weighting.
• Device and session de-duplication to stop one passenger inflating a cut.
• Screen health monitoring: plays from screens with GPS or uptime faults are excluded from measurement, not just flagged.
• Quarterly calibration against an external population survey to detect drift.
• Every published figure carries a sample size and confidence interval in the UI and the export.

Operating model

• Data Services is a separate P&L reporting to the CEO, not a cost centre inside media.
• Media sells the measurement attach; Data Services delivers it and books the revenue.
• A monthly Data Council, CEO, CTO, Head of Data, COO, signs off on methodology changes and any new published metric.
• Every campaign has a named analyst; no automated report leaves the building unreviewed in H1.