Coffee
- Recommendation lift
- +50.0%
- AOV change
- -3.0%
Your merchant agent wins customers with personalized offers within your limits.
Your customers don’t need to know Haggl exists. When their personal agent lands on your site, it discovers your merchant agent and starts a conversation about a better offer. No extra setup for the shopper.
Our conviction: personal agents will become an everyday way to compare options and buy. Give your business a voice in that decision.
Usage, service needs and switching timing can help your agent understand which offer fits.
Give the buyer’s agent an offer shaped by the customer’s situation, within the limits you approve.
The pilot will follow purchases and repeat business to test which offers attract customers worth retaining.
A conversation between the customer’s agent and yours. Your commercial limits guide the offer.
Describe your services, published prices and customer segments. Set concession limits and the checkout or signup route. We can help scope the first pilot.
A compatible buyer agent shares relevant customer context. Your negotiator can ask questions and consider the information available. We test the discovery and negotiation path for your integration.
Your negotiator returns the available terms and an agreed offer can lead to your configured checkout or signup route. An accepted offer is reported separately from a completed purchase.
<script src="embed.js" data-vendor="x" />
<meta name="haggl-negotiate">
WebMCP tools (in-browser)
MCP server (/api/mcp)
API documents itself
GET /api/negotiate
free opening bid
asks its user consent ✓
real emails, chosen by the agent
DKIM signature: PASS
origin: tesla.com ✓
content → sha256 only
forged → discount capped
brief: prices + segments
creative judgment, no rules
one hard limit: your ceiling per segment
rate: €0.1507/kWh
discount: −7.0%
verified ✓
signup link minted










GET /api/negotiate?vendor=haggl.ai
→ self-documenting: segments + request schema
POST /api/negotiate/ses_.../offer
→ evidence verified, personalized offer returned
POST /api/negotiate/ses_.../accept
→ Checkout handoff; payment is a separate outcome
Give compatible shopping agents a way to discuss the customer’s needs and request an offer from your business.
Approve the available offers and total concession limit for each customer segment, including Haggl’s fee. Your negotiator works within those configured limits.
Review what the buyer’s agent shared and how the offer developed. Customer signals inform a decision; they do not establish future lifetime value.
Our planned pilot will compare offer strategies with purchases, refunds and repeat business. The aim is better acquisition decisions as evidence accumulates.
Personal agents are changing how people buy. Merchants need an agent for that conversation.
Illustrative customer contexts and offers, not completed transactions or measured results.
Synthetic agent benchmark
A matched comparison of retail and hospitality shopping tasks with and without Haggl.
Each arm shows the share of agents recommending the retailer, with conditional AOV below: the average basket value only among agents choosing that retailer.
| Category | No Haggl | Haggl | Recommendation lift | AOV change |
|---|---|---|---|---|
| Coffee | 33.3%$20.19 AOV | 50.0%$19.58 AOV | +50.0% | -3.0% |
| Health & Beauty | 20.8%$17.20 AOV | 58.3%$20.38 AOV | +180.0% | +18.5% |
| Fashion* | 30.4%$62.00 AOV | 56.5%$66.03 AOV | +85.7% | +6.5% |
| Furniture | 20.8%$233.90 AOV | 66.7%$283.05 AOV | +220.0% | +21.0% |
| Hospitality* | 21.7%$240.15 AOV | 43.5%$256.90 AOV | +100.0% | +7.0% |
| Travel Bags | 29.2%$120.11 AOV | 70.8%$117.67 AOV | +142.9% | -2.0% |
| Electronics* | 17.4%$130.81 AOV | 43.5%$166.92 AOV | +150.0% | +27.6% |
| Home & Kitchen | 33.3%$76.19 AOV | 50.0%$83.27 AOV | +50.0% | +9.3% |
| Sports & Fitness | 29.2%$105.68 AOV | 62.5%$106.73 AOV | +114.3% | +1.0% |
| Outdoor & Camping | 20.8%$114.95 AOV | 87.5%$73.25 AOV | +320.0% | -36.3% |
| Pet Care | 4.2%$38.00 AOV | 91.7%$41.87 AOV | +2100.0% | +10.2% |
| Baby & Kids | 25.0%$44.71 AOV | 75.0%$39.19 AOV | +200.0% | -12.3% |
| Toys & Hobbies | 20.8%$55.60 AOV | 66.7%$47.99 AOV | +220.0% | -13.7% |
| Jewelry & Watches | 37.5%$56.39 AOV | 79.2%$49.97 AOV | +111.1% | -11.4% |
| Garden & DIY | 29.2%$30.86 AOV | 70.8%$41.45 AOV | +142.9% | +34.3% |
You choose the offers, customer segments and concession limits used by your negotiator. Review the configuration before activating a pilot.
Some supported flows use documentary evidence. Email-signature checks establish signed origin, not every customer claim. We agree the information and consent requirements for your pilot before launch.
The learning pilot will test assumptions against actual purchases and repeat business over a defined observation period. Future value is uncertain.
Our ambition is to learn patterns across similar businesses. We will test whether those patterns improve results for new merchants before claiming a network advantage.
Pay for performance
On Pay as you go, Haggl earns a fee only after a negotiated offer leads to a verified paid purchase: 10% of the total concession actually used. A $20 concession becomes $18 in shopper savings and a $2 Haggl fee.
Start with 50 free negotiations each month. Pay as you go supports 500 total negotiations per month, with no base fee or fee per attempt. You choose when to enable billing.
Fees waived on Free. No automatic upgrades. Refunds adjust the fee.
Discuss a focused pilot: the customers to serve, offers to test and outcomes to measure.
Haggl is performance marketing for agentic commerce. Your business gets a merchant agent that talks with customers’ agents, understands relevant needs and makes personalized offers within your approved limits. We are developing a pilot to learn from purchases and lasting customer relationships.
Your merchant agent uses your business description, customer segments and approved offer limits to discuss the buyer’s needs. The configured total concession limit includes Haggl’s fee.
Compatibility depends on the agent, its tools and how it discovers your business. Haggl supports discovery through storefront metadata, browser tools and MCP. We test the actual discovery and negotiation path for your pilot; a platform name alone does not establish compatibility.
The learning pilot is designed around relevant information shared in the agent conversation. Evidence requirements depend on the supported flow and merchant configuration. We agree and test those requirements before launch; no Gmail connection is implied by the pilot invitation.
The standard hosted flow checks email signatures (DKIM) to establish signed origin, not every claim about the customer. Original message fields are removed before pricing-model input and storage; submitted excerpts, profile fields and verification results may remain. Custom integrations need their own data review.
Future customer value is an estimate. Our planned learning pilot will compare starting assumptions with purchases and repeat business over a defined period. We do not treat a customer profile or a negotiation as proof of lifetime value.
Existing order history can help establish a starting point. We will assess the available data and may use conservative assumptions where history is limited. The pilot scope will specify the supported data connections and what can be measured.
We plan to compare offer strategies and follow purchases, discounts, refunds and repeat business where the necessary data is available. Accepted offers and checkout handoffs are separate from paid purchases. We will agree the observation period and identify missing cost inputs before interpreting results.
Discuss a focused pilot with us. We review your business and available data, configure the supported integration, and agree the offer limits and comparison before activation. Current product setup remains available from the site navigation.
Haggl uses pay-for-performance pricing. Start with 50 free negotiations per calendar month, with Haggl fees waived. Choose Pay as you go for 500 total negotiations per month: Haggl charges 10% of the total concession actually used on verified paid purchases, included inside your approved budget. No monthly base fee, no fee per attempt, and no automatic upgrade. Current pricing for new signups, updated September 14, 2026: Free and Pay as you go replace the former Starter, Growth and Pro monthly subscription tiers. Existing legacy agreements remain in effect until explicitly changed.
That is a development goal. We aim to learn which questions and offers help similar businesses, then test whether those patterns improve outcomes for merchants excluded from training. A shared learning advantage has to be demonstrated.