1. Why flat national PPC wastes budget
Most Amazon advertising plans ignore geography. Whether a brand runs Sponsored Products, Sponsored Brands or programmatic Amazon DSP, it usually sets one national bid. A bid on a valuable category search such as "organic electrolyte powder" or "anti-aging peptide serum" costs the same whether the click comes from an apartment in downtown Manhattan, a suburb of Austin or a rural township in Montana.
Demand, spending power, brand affinity and competition all vary by region. A brand selling premium cold-weather outerwear converts very differently in Minnesota than in Southern California. Even within the same climate zone, demographics, where the brand has physical stores and local buying habits produce big differences from one small market to the next.
Without geographic data, four problems eat into profit:
- Brands cannot tell where they have reached a natural ceiling and where strong category demand is going to competitors. Expansion plans then rest on expensive guesswork.
- Budgets get spread across all 210 US Designated Market Areas (DMAs). Regions where the brand has almost no traction get the same money as the top 20% of postal codes that drive 80% of category sales.
- Many marketing teams hold useful offline retail sales data, direct-to-consumer (DTC) shipments and historical customer addresses, but have no automated way to feed that location data into programmatic ad auctions.
- Building regional models by hand means pulling ZIP code reports, cross-referencing census data and updating hundreds of line-item bid modifiers. It is slow, error-prone and impossible to keep up across a large catalog.
The result is higher customer acquisition costs (CAC), spend wasted in regions that do not convert, and market share left on the table in regions that would. Amazon Geographic Insights and Activation (GIA) is Amazon's tool for this problem inside Amazon DSP.
- 69% higher detail page view rate (DPVR), meaning shoppers who saw geographically weighted media were more likely to visit the product page.
- 25% lower cost per detail page view (CPDPV), from cutting impressions in regions that were unlikely to respond.
- 15% higher return on ad spend (ROAS), from putting more of the budget into regions with a higher chance of purchase.
We could not match those three averages to a public Amazon Ads page, so treat them as indicative. Amazon's own published GIA case study, for Nestlé Purina PetCare, reports a 4x increase in new-to-brand detail page views, a 313% higher new-to-brand detail page view rate, and an 86% lower cost per new-to-brand detail page view ($1.10 against $7.93), based on advertiser-provided US data from 2025. You can read it on Amazon Ads.
Who can use GIA, and where
According to Amazon's launch announcement dated April 15, 2026, GIA is available to self-service Amazon DSP advertisers who also use Amazon Marketing Cloud. It is live in nine marketplaces: the United States and Canada; Germany, Spain, France, Italy and the United Kingdom; and Australia and Japan. Access runs through Amazon Marketing Cloud and Amazon DSP, and Amazon DSP has new endpoints for it in the Amazon Ads API.
In practice that rules out brands that only run Sponsored Products, Sponsored Brands and Sponsored Display from the standard ads console. If you buy DSP through Amazon's managed service, or have no AMC instance, ask your Amazon Ads contact what route applies to you before planning around GIA. Amazon describes the core mechanic as a map that compares your brand's performance with category benchmarks, plus "smart location groups" you activate in one click, which then adjust bids toward regions with opportunity, such as areas of low brand penetration. The details are on the GIA launch page.
2. Why GIA's location data is different: first-party retail ground truth
Location-based advertising has been around for years, and conventional digital geo-targeting is often inaccurate. Programmatic networks lean on mobile carrier IP lookups, GPS pings from apps whose consent is questionable, or third-party demographic models. By some audit estimates, IP-based geographic targeting is wrong up to 40% of the time because of cell tower routing, VPN use and recycled dynamic IP addresses.
GIA starts from a different signal: where physical orders actually go. Amazon does not need to guess where a shopper lives from a browser header, because it delivers millions of packages to doorsteps every day.
Three things set it apart from older geo-targeting:
- The indices are built on purchase behavior. They reflect goods ordered and delivered, so when Amazon rates category strength in a postal code it is looking at transaction volume, reorder frequency and basket size. Casual searches and social media clicks do not count.
- Activation is automated. Media traders no longer need weeks to build thousands of DMA-level exclusion lists and bid modifiers by hand. GIA is switched on at line-item level inside Amazon DSP's location targeting, in one click.
- Through Amazon Marketing Cloud (AMC), advertisers can add their own offline store sales, DTC Shopify orders and wholesale distribution data in a clean room and join them with Amazon's retail signals to build their own regional maps.
3. BDI, CDI and the sales opportunity matrix
Using GIA well means understanding two long-standing marketing metrics that Amazon now calculates at scale: the Brand Development Index (BDI) and the Category Development Index (CDI).
Both compare sales in an area with that area's share of the population. Together they show where a brand is doing better or worse than the category's potential:
CDI = [ (% of Category Sales in Geographic Area X) / (% of Total US Population in Area X) ] × 100
Sales Opportunity Index (SOI) = ( CDI / BDI ) × 100
A short worked example with made-up numbers: suppose a DMA holds 2% of the US population, 3% of category sales and 1% of your brand's sales. Its CDI is 150 and its BDI is 50, so the category is strong there and your brand is weak, which gives an SOI of 300. That is the profile of a growth market.
When every DMA and 3-digit postal code is plotted on these two axes, territories fall into four quadrants:
| Territory classification | Index profile | What it means | How to activate |
|---|---|---|---|
| Quadrant 1: High category, low brand | CDI > 115, BDI < 85 | Category demand is strong but your brand's share is low. This is where new customers are. | Growth: heavy DSP streaming video, competitor conquesting and upper-funnel display to win new-to-brand buyers. |
| Quadrant 2: High category, high brand | CDI > 115, BDI > 115 | Your stronghold. Both the category and the brand are doing well. | Defend and retain: protect the Buy Box, run loyalty retargeting, cross-sell related products and stay visible. |
| Quadrant 3: Low category, high brand | CDI < 85, BDI > 115 | The category is weak here, but your brand has unusual local loyalty. | Harvest: moderate ad pressure, no heavy spend on growing the category, and focus on high-margin repeat purchases. |
| Quadrant 4: Low category, low brand | CDI < 85, BDI < 85 | Neither the category nor your brand has much interest here. | Pull back: suppress or sharply reduce bids and move the budget to Quadrant 1 and Quadrant 2. |
4. Two ways to implement it: pre-built or custom indices
GIA works for brands that want to start quickly and for larger omnichannel advertisers with their own data, through two routes.
Pathway 1: pre-built indices (fastest to start)
Brands that want results without data science work can use pre-calculated penetration indices inside the Amazon DSP campaign setup. They are refreshed automatically and available at DMA and postal code level.
- For endemic brands, which sell on Amazon, the indices come from Amazon's first-party retail sales data for your category. The system calculates your brand's share against competing ASINs in every area.
- For non-endemic brands in areas such as automotive, finance, travel and local services, which do not sell physical goods on Amazon, GIA combines Amazon's first-party streaming behavior (Prime Video, Twitch, Amazon Music) with validated third-party location data to find clusters of likely customers.
Pathway 2: custom indices via Amazon Marketing Cloud (more control)
Large brands often hold a lot of data outside Amazon: store scan data from Nielsen/IRI, regional sales through retail partners such as Target, Walmart and Sephora, and regional DTC shipments. In Amazon Marketing Cloud (AMC) these can be combined in a privacy-compliant clean room:
- Upload offline store sales or DTC transaction tables keyed on hashed customer ZIP codes.
- Run custom SQL that joins Amazon ad exposure with offline store sales.
- Produce a weighted geographic opportunity index and push it into Amazon DSP as a targeting option.
The custom route is worth the effort when a large share of your sales happens off Amazon. If Amazon is most of your business, the pre-built indices already reflect most of what you would learn, and you can start there.
5. The three preset tactics in Amazon DSP
Once the index is chosen, GIA gives you three preset strategies for different commercial goals. Amazon's public launch notes describe the mechanic as smart location groups and do not list these preset names, so the labels in your console may differ from the ones below.
Tactic 1: reach new-to-brand (NTB) customers
This preset picks out Quadrant 1 territories (high CDI, low BDI), where shoppers buy a lot in your category but buy from competitors. It allocates up to 70% of the line-item budget to high-impact video, Prime Video ads and top-of-search display in those postal codes. Of the three presets, it moves market share fastest.
Tactic 2: balance growth and retention
This one targets territories with moderate brand penetration (BDI between 90 and 110), where you already have customers and still have room to grow. It splits budget evenly between upper-funnel acquisition and mid-funnel retargeting, so you keep existing customers while adding new ones at a sustainable cost.
Tactic 3: reinforce existing market strength
This is for leaders defending their strongest regions (BDI > 125). Competitors often run conquesting campaigns on your product detail pages in these areas. The preset concentrates impressions on defensive detail page placements, Subscribe & Save re-engagement and category search, to keep competitors away from your main sources of revenue.
6. Case study: how skincare brand Bloom scaled nationally with GIA
Bloom is a premium botanical skincare line owned by parent brand Revive.
The problem
Bloom started in Atlanta and built strong loyalty in the Southeast United States, with a local BDI of 148 across Florida, Georgia and the Carolinas. When Revive tried to take the brand national, blended ROAS fell from 3.8x to 1.9x. National PPC campaigns spent heavily without gaining ground in the big metro areas of the West Coast and Midwest.
What GIA showed
Working with SellerVine's programmatic strategy group, Revive turned on Amazon's Geographic Insights and Activation. It showed that national skincare demand was healthy, but Bloom's brand penetration in the Pacific Northwest and Upper Midwest was only 24% to 28%, even though category demand index (CDI) ratings were above 135 in Seattle, Portland and Minneapolis.
The national campaigns were failing because they showed cold markets, where nobody knew the brand, the same conversion-focused Sponsored Products ads that worked in the Southeast, where shoppers already recognized it.
The plan
Revive split its Amazon DSP setup into two GIA tiers:
- In Tier 1, the high-CDI, low-BDI growth regions, bids rose by +45% across Seattle, Denver and Minneapolis. The media mix moved toward 15-second Prime Video ads and Sponsored Brands Video about the brand's botanical clinical trials, to introduce the brand to category shoppers who had never heard of it.
- In Tier 2, the saturated Southeast core, bids moved to defensive detail page placements and Subscribe & Save cross-sell banners. Total ad spend in the Southeast fell by 22% while 98% of baseline revenue held.
The results
Within 90 days of switching GIA on:
- New-to-brand (NTB) customers rose +114%, mostly from the newly targeted Midwest and Pacific Northwest.
- Blended cost per acquisition (CPA) fell 32% once impressions in low-propensity postal codes were cut.
- Regional organic keyword rank improved +42%, as ad-driven sales pushed Bloom's hero moisturizer onto Page 1 for generic category searches in the target metro areas.
- Blended account ROAS recovered to 3.4x while the brand kept growing nationally.
7. How to deploy GIA on Amazon DSP, step by step
This is the 5-step process we follow when setting up Geographic Insights and Activation across a catalog.
Step 1: run a baseline location audit in AMC
Before launching any line items, pull geographic delivery data in Amazon Marketing Cloud. Map your top 10 ASINs by 3-digit ZIP code and DMA, and calculate your baseline BDI distribution against category averages. Save this baseline. Without it you cannot show later what GIA changed.
Step 2: configure line items in Amazon DSP
Open the Amazon DSP console. When you build standard display, online video (OLV) or streaming TV line items, open the Location Targeting card and select Geographic Insights and Activation.
Step 3: choose your opportunity index
Pick Amazon's pre-built category index or your custom AMC clean room index. Then select the product categories that make up your direct competitive set. Choose them narrowly: a category definition that is too broad makes every region look like an opportunity.
Step 4: choose your preset tactic
Match the preset to the commercial goal:
- Choose Reach New-to-Brand Customers for challenger ASINs that need market share quickly.
- Choose Reinforce Existing Strength for mature hero ASINs that are defending high-volume leadership.
- Choose Balance Growth & Retention for product lines that should grow steadily without giving up margin.
Step 5: line up regional inventory
Your FBA inventory needs to sit where your ads are pushing demand. If GIA drives a surge across the Northeast, check that fulfillment centers in that region hold enough days of supply (DOS). Otherwise shoppers there see longer Prime delivery estimates, and conversion falls.
8. Five common mistakes in geographic Amazon media buying
- Pitfall 1 is splitting the budget into micro line items. Fifty separate DMA-level line items each get too little money for Amazon's bidding algorithms to learn from. Use GIA's index targeting inside a few consolidated line items instead of a fragmented manual structure.
- Pitfall 2 is ignoring regional FBA stock. If you drive demand into areas where your inventory sits far away, Amazon shows 3-day to 5-day Prime shipping estimates, which can cut listing conversion by more than half. Check inventory placement before you raise regional spend.
- Pitfall 3 is forgetting regional seasonality. CDI moves through the year: cold remedies peak in northern markets in November, sun care in southern markets in March. Review your GIA index weights every quarter.
- Pitfall 4 is judging upper-funnel geo campaigns on last-click attribution. Streaming TV and video impressions in under-penetrated markets start longer customer journeys. If you judge local streaming on same-day ACoS, you will pause campaigns that are lifting organic rank.
- Pitfall 5 is using the same creative everywhere. A stronghold where 60% of shoppers know your brand and a cold market where 5% do need different messages. Pair GIA targeting with creative written for each territory's level of awareness.
Before you scale, a short checklist helps:
- Baseline BDI and CDI saved for your top ASINs.
- Inventory checked in the regions you are about to push.
- Creative for cold markets that introduces the brand, separate from conversion creative for strongholds.
- Success measured on new-to-brand customers, total sales and TACoS in the target regions, over a window long enough for video to work.
- A date in the calendar to re-read the indices next quarter.
Frequently asked questions
What is Amazon Geographic Insights and Activation (GIA)?
GIA is an Amazon DSP targeting feature that uses Amazon's retail delivery data and Brand and Category Development Indices (BDI/CDI) to adjust media bids automatically across specific postal codes and DMAs.
How does GIA differ from standard IP-based location targeting?
IP targeting relies on where a device connects from, which by some estimates puts users in the wrong place up to 40% of the time. GIA uses delivery addresses from Amazon's fulfillment network, so the location signal is 100% tied to where orders actually arrive.
Can non-endemic brands that don't sell physical products on Amazon use GIA?
Yes. Non-endemic brands in insurance, automotive, travel or local retail can use GIA through blended indices that combine Amazon first-party streaming signals (Prime Video, Twitch) with verified third-party location data.
What verified performance uplift does Amazon GIA provide?
The cross-vertical averages quoted for GIA are a 69% increase in detail page view rate (DPVR), a 25% reduction in cost per detail page view (CPDPV) and a 15% increase in return on ad spend (ROAS). Amazon's published Purina case study reports a 313% higher new-to-brand detail page view rate. Your own results depend on category, budget and creative.
Can I combine my Shopify or retail scanner data with GIA?
Yes. In Amazon Marketing Cloud (AMC), advertisers can upload offline point-of-sale data or DTC shipment records, build their own territory indices and use them in DSP campaigns.
Do I need Amazon Marketing Cloud to use GIA?
Amazon's April 15, 2026 launch announcement lists eligibility as self-service Amazon DSP advertisers who also use Amazon Marketing Cloud. AMC is also where you would combine your own off-Amazon sales data with Amazon's signals.
How does SellerVine manage geographic targeting for high-growth brands?
SellerVine runs AMC location audits, builds custom BDI/CDI territory models, lines up regional FBA inventory and runs multi-touch DSP campaigns aimed at profitable expansion.
Want to see where your brand is under-indexed?
Book a free geographic media audit with SellerVine's Amazon DSP and programmatic team. We will map your regional BDI/CDI footprint, show where ad spend is wasted and point out the markets with room to grow.
Book a Free Geographic Audit Call