Shoppable Screensavers on DirecTV Gemini in 2026: STC KPI Playbook

Screensavers just became storefronts

DirecTV and Glance AI have announced a partnership to bring AI-powered, shoppable screensavers to DirecTV’s Gemini devices starting in early 2026. The idea is simple and far reaching: when a TV goes idle, the screensaver shifts from static wallpaper to an interactive surface that features personalized scenes and products viewers can buy with a phone scan or tap. For growth and marketing leaders, this is not about another ad placement, it is a new opt-in format that stitches the lean-back TV moment to a lean-forward mobile checkout.

In the announcement, the companies emphasized that Gemini devices will host interactive, AI-driven experiences that connect content discovery and commerce. For confirmation direct from the source, see the official release on Glance AI coming to Gemini. That timeline gives teams roughly four quarters to test the TV-to-phone handoff, design consent flows, and define a new conversion metric that starts at the screensaver. Coverage has also highlighted the ability to generate AI scenes and then buy similar items via phone, see reporting on AI-generated, shoppable screensavers.

What exactly is screensaver commerce

Screensaver commerce turns idle-screen time into a personalized shopping canvas. Instead of a generic slideshow, viewers see scenes that can include their preferences and, if they opt in, avatar-style representations of themselves or their households. The scene surfaces shoppable items that match the setting. A QR code or on-screen prompt hands the session to a phone for product detail, customization, and payment. The purchase completes off-device, typically in a mobile browser or app controlled by the brand or retailer.

Early previews point to interaction patterns that matter to growth teams: avatar creation and management inside a companion app, voice or remote-based controls to adjust scenes, and product suggestions that mirror what appears on screen. The core growth problem becomes obvious: how do you reduce friction in the handoff from a passive, communal TV moment to a personal, authenticated mobile checkout.

The KPI you will wish you had named today

Most teams will reflexively measure QR scan rate and click-through rate. Good start. The metric that will matter more is screensaver-to-cart, abbreviated STC. Define it precisely now so pilots and 2026 launches roll into a single optimization graph. For adjacent KPI design, see how to add agent conversion KPIs.

• Working definition: STC is the percentage of TV screensaver impressions that lead to an attributed cart creation on a linked mobile device within a set window, typically 30 minutes.
• Why it matters: STC captures intent that survives a large context shift, communal TV to personal phone. It rewards creative and engine tuning that lowers cognitive load, not just CTR bait.
• How to instrument: tag the TV experience with a session ID embedded in the QR code URL or deep link, pass that ID through your mobile web or app, and fire a cart-created event with the same ID. Stitch in your CDP or analytics layer. If you cannot stitch deterministically, create a probabilistic fallback that uses time, device model, IP block, and first page path, then discount by a calibrated factor.

Once you define STC, you can ladder additional KPIs underneath it: opt-in rate to avatar creation, scene personalization completion rate, device-to-phone handoff latency, product detail page dwell time, and purchase within 24 hours. Together these give you a clean funnel from idle scene to checkout.

Pilot design, a 90-day blueprint

Use the remainder of 2025 to run a structured pilot in three phases so you enter 2026 with data, not opinions. Borrow checkout acceleration tactics from our 90-day chat commerce playbook.

Phase 1, feasibility and consent, weeks 1 to 4

• Audience: 5 to 10 percent of your addressable base that also overlaps with DirecTV Gemini households. If you do not know that overlap, run a voluntary signup list and recruit via email or SMS.
• Consent: make the experience strictly opt in. Ship a consent screen that describes avatar creation, data usage, and how to delete assets. Include a one-click forget me control.
• Creative: launch three scene templates aligned to product categories with clear visual anchors: apparel try-ons, living room furniture, and outdoor gear.
• Handoff: support both QR and typed short URL with code entry for viewers who prefer not to scan. Track QR scans and code entries separately.

Phase 2, optimization and targeting, weeks 5 to 8

• Targeting: test contextual rules tied to TV time-of-day and household signals. For example, weekday evening scenes emphasize apparel bundles, weekend midday scenes emphasize home decor sets.
• Offer design: test no-offer versus small value adds like free shipping. Watch for offer selection bias inflating STC. Keep offers consistent across cohorts.
• Latency budget: set a sub-2 second load time target from scan to first paint on the mobile PDP. Anything slower will crater completion rates.

Phase 3, scale and attribution, weeks 9 to 12

• Scale: expand to 20 to 30 percent of the pilot panel. Guard against frequency burn, cap exposures to three unique scenes per day per household.
• Attribution: use session stitching for last touch and a 24-hour halo model for view-through on TV-to-phone paths without scans. Refine with a geo-split or holdout if you have enough coverage.
• Decision rule: greenlight wider investment if STC lifts total cart creations by at least 8 to 12 percent for the pilot cohort with neutral net contribution margin.

Creative that reduces handoff friction

TV is a shared surface. Phone is personal. Your creative must explain the bridge in a single beat.

• Use a persistent visual affordance: a QR code block with a short, pronounceable URL. Include a three-word verb phrase like Open my look or Shop this room.
• Keep the scene simple: one to three focal products, clear color and texture, no cluttered backgrounds.
• Show the mobile endpoint: place a small over-the-shoulder phone mockup in the corner with the first screen of the journey, so viewers know what they will land on.
• Build bundles: pre-assemble two or three products that appear together in the scene and land users on a prefilled cart, not a lonely PDP.
• Localize subtly: use weather or local sports motifs that fit the day without creeping out households.

Consent, privacy, and brand safety for AI avatars

If you plan to show AI-rendered people or personalized avatars, craft a policy before you ship your first test. Earn, do not assume, consent. For a broader privacy approach, review our privacy-safe chat signals playbook.

• Consent tiers: separate three levels: generic scenes with no people, scenes with non-identifiable silhouettes, and personalized avatars. Each tier needs its own opt-in, its own retention policy, and a clear opt-out path.
• Asset lifecycle: store source photos and generated assets with separate keys, add a time-to-live, and set deletion to run within 24 hours of an opt-out. Log suppressions to prevent reactivation from downstream caches.
• Household rules: if a TV serves multiple accounts, bind avatars to a personal mobile account and never persist them on the TV device. Keep the TV experience stateless and privacy-preserving.
• Safety filters: block sensitive categories by default, for example kids apparel try-ons, health devices, or anything that implies protected attributes. Add a manual review queue for edge cases before a template goes live.
• Transparency: watermark all AI scenes with a small AI generated label and add a tap-through explainer on mobile.

Engineering the TV-to-phone bridge

The handoff is your make-or-break moment. Treat it like a checkout page.

• Link structure: use universal links on iOS and Android App Links so scans open your app when installed and fall back to mobile web when not. Carry a signed session token to attribute STC cleanly.
• Device detection: prefetch images and the first API call based on common device and network conditions. Show a skeleton layout if the first content cannot arrive in 500 milliseconds.
• Payment: enable express checkout with wallets and stored profiles. Offer a guest route that does not force account creation before cart commit.
• Accessibility: size the QR code square to at least 180 pixels on a 55 inch TV at 8 to 10 feet viewing distance. Include high contrast borders and a 6 to 8 character short code for manual entry.

Measurement, the model that will survive finance reviews

Define a simple, defensible attribution approach that your finance partners will trust.

• Deterministic path: scans or short-code entries that carry a token to mobile are last touch for STC and cart creation.
• Probabilistic view-through: impressions without scans that correlate to uplift in cart creation within 24 hours get a fractional credit, for example 0.2, based on historical calibration from geo-splits.
• Clean room option: if you have retailer partners, use a neutral clean room to match session IDs to purchase events with privacy controls. Keep your contribution analysis on your side, share only aggregated deltas.
• Guardrails: forbid cross-household stitching, disallow IP-only matching, and force minimum cohort sizes for any reported metric.

Where to start, categories and audiences

Not every category fits a screensaver. Lean into visual-first items with low configuration cost.

• Apparel and accessories: high match to avatar scenes, simple bundling.
• Home decor and small furniture: clear scene-to-product mapping, natural bundling.
• Beauty and grooming: consider sampler bundles that ship as a set.
• Sports and outdoor: seasonal rotations with equipment and apparel together.

Start with mid-funnel audiences that already engage with your brand on mobile. Avoid cold starts until you validate STC above your baseline cart creation rate by at least 5 points.

Partner and platform considerations

Ask specific questions of platform partners and internal teams.

• Gemini device footprint: confirm install base in your target geos and the ratio of Gemini to total DirecTV households in your overlap.
• Controls: verify opt-in defaults, frequency caps, and quiet hours where the screensaver will not trigger shoppable scenes.
• Data boundaries: ask where on-device processing ends and what, if any, event-level data flows to third parties. Insist on on-device generation for avatars and strictly scoped event export.
• Retailer links: ensure retailers can ingest session tokens from your links so you still attribute STC when the cart lives on a partner property.

Creative operations and governance

Treat screensaver commerce as a distinct surface in your content calendar.

• Templates: maintain a small library of approved scene templates with known brand-safety properties. Ban new templates without review.
• Copy rules: write microcopy for the QR prompt that matches your privacy posture and tone. Keep it factual and clear.
• Review cadence: schedule weekly reviews on performance and incident checks, for example false personalization, misgendered avatars, or product mislabeling.
• Incident response: document a takedown process that can suppress a creative in less than 30 minutes with a rollback link ID list.

Teams using Upcite.ai often keep a single shoppable idle-screen brief that includes consent language, data retention rules, and acceptance criteria for STC lift. That keeps marketers, legal, and engineering aligned on what good looks like before a scene ships.

Risks and the mitigations that matter

• Creepy factor: viewers do not want surprise personalization. Keep default scenes generic, then invite opt in to personalization with clear value, faster outfit building or room design.
• Household mix: do not assume one taste fits all. Make scenes session based, not persistent. If someone else turns on the TV, show neutral content.
• Measurement fog: without clean handoff you will overcount. Force a tokenized handoff for any scan, then strip analytics for impressions without scans to a view-through model with discounting.
• Retail out-of-stocks: AI scene recommendations must check inventory before showing products. Nothing kills STC like an unavailable look.

Budgeting, what to expect in year one

A pragmatic first year plan looks like 2 to 5 percent of upper funnel digital spend diverted into screensaver commerce tests, with guardrails on contribution margin. Expect early STC in the 0.3 to 0.8 percent range depending on category, with leaders cracking 1 percent by quarter three as creative and latency improve. Treat any CPA comparisons carefully since this format drives bundle purchases and cross-sells that are not cleanly comparable to single SKU ads.

Legal and policy checklist

• Consent logs: stored for at least one year with immutable records.
• DPA updates: reflect avatar assets and deletion SLAs.
• Content policy for minors: no personalized avatars for accounts registered to users under 16. Apply a no-people template set for kids households.
• Disclosure: small but legible AI disclosure on both TV scene and mobile landing.
• Trademark hygiene: if your scene shows brand-like aesthetics, confirm no implied endorsement unless you have rights.

The 2026 clock is running, next steps

This rollout starts in early 2026 on Gemini devices, which means your window to prepare is now. Here is a simple action plan you can execute in the next 8 weeks.

1. Stand up measurement: define STC and build the tokenized handoff with your analytics team. Add supporting events: opt-in rate, scene completion, cart creation, and 24 hour purchase.
2. Build consent and safety: write your avatar consent tiers, asset deletion rules, and brand-safety filters. Get legal sign off.
3. Choose categories: select two visual-first categories and draft three scene templates per category. Prepare bundles and prefilled carts.
4. Run a 90-day pilot: recruit a panel, test creative and offers, and set a greenlight rule tied to STC and contribution margin.
5. Prepare scale: codify frequency caps, schedule creative governance, and coordinate with platform partners on device footprint and controls.

This is not a gimmick. It is a new, measurable path from living room attention to mobile carts. Teams that pilot now with an opt-in mindset, a screensaver-to-cart KPI, and a clear consent playbook will be ready to scale when Gemini screens light up in 2026.