Postal
Postal is a corporate gifting and direct mail platform for revenue and people teams. Customers send physical gifts, branded swag, electronic gift cards, direct mail and virtual experiences from a curated marketplace, either one to one or through a shareable link carrying a single item or a collection, and can trigger sends automatically from events in a connected record system such as a renewal or a work anniversary. Warehousing, pick and pack, international fulfilment and a company branded storefront sit alongside the marketplace, and an in house creative agency designs custom items.
The mechanism a buyer should understand is the acceptance step. Recipients are notified by email or by link that a gift is waiting and enter their own shipping address to accept it, so the sender never needs to hold a home address and nothing ships to a stale one. The vendor states it charges only for gifts a recipient accepts, which is the largest cost in this category.
Sendoso acquired Postal on 30 April 2025. The product still sells under its own name on its own domain, with a free single user tier reached from the site navigation and a quoted paid tier above it, and the homepage carries a banner announcing the acquisition. The integration is partway done and it shows in the legal footer: the subprocessor list is hosted on the acquirer's domain, while the footer offers two separate terms documents whose addresses differ, the primary one marked as superseded in its own address. A buyer should establish which document governs before signing.
No price is published at any level. The pricing page shows one plan, a feature list and a route to a sales conversation, and does not mention the free tier that the site navigation offers.
Capability Axes
Capability grades
17 of 17 axes rated · 4 graded A or B
The word intelligent carries the positioning and nothing on the retrieved pages carries the word. The vendor calls itself the leading intelligent gifting platform in its page title, its meta description and its homepage headline, and across the homepage, the pricing page and both question sets there is no model, no generated copy, no recommendation engine and no scoring feature described.
Strip the adjective and the product is unchanged: a marketplace, a fulfilment network, warehousing, a storefront builder and trigger automation, all of which sell on their own. That is a complete product rather than a veneer, which is why this sits at the middle of the band and not at the floor, but the adjective is doing work no described capability supports. The acquirer has shipped model driven gift recommendation on its own platform and none of it appears on this one. Ask which functions on this platform are model driven and which are rules.
This system spends money without a person in the loop, and what bounds it is a funded balance. Sends can be triggered automatically from events in a connected record system, so a rule rather than a human decides that value leaves the account, which is the highest consequence form this axis takes in any category. Two controls are published against that.
Spending draws on a prepaid balance funded by card or invoice, so a misconfigured rule cannot spend past what has been loaded, and budgets can be set and funds pre allocated per user. Nothing published addresses approval steps before a triggered send, per campaign or per recipient value caps, or who may create an automation. The bound is therefore financial rather than procedural, and it stops a runaway only after the balance is gone. Ask what approval and value limits can be enforced on an automated send, and who can change them.
There is no model layer disclosed and none found. No provider, model family or version is named anywhere retrieved, no content is described as generated, and there is no scoring, ranking or routing logic whose explainability a buyer could ask about. The nearest thing to automated decisioning is trigger automation from record system events, which is rule based and configured by the customer rather than inferred.
Graded at the middle of the band with the scope stated rather than at the floor, because the floor asserts that an undisclosed model is operating and that is not what was established here. The gap that matters for a buyer is forward looking: the parent company has announced model driven gift recommendation trained on a combined gifting database, and nothing on this product's surface states whether that reaches this platform. Ask whether any model runs against your account data today, and which one.
Better than a logo wall and short of a measurement. Testimonials are attributed to named individuals with their job titles and employers rather than floating free, which is a real evidence surface, and named customers include several large technology companies.
One quantified figure appears, a claim that forty percent of leads receiving an incentive through the platform convert to pipeline, and it is a customer's assertion inside a testimonial rather than a vendor measurement: no sample size, no period, no comparison group and no method accompany it. A third party ranking badge from 2024 and review platform badges are displayed.
The pricing page carries review badges dated to the summer of 2023 while the homepage carries badges dated to 2026, so one live page advertises standing three years out of date. Ask for acceptance rate and cost per accepted gift across a comparable customer set, with the period stated.
An acceptable use policy is published as its own instrument, which is more than most of this category offers, and the regulated question this product actually raises is untouched. The vendor does not operate the outreach channel: shareable gift links are dropped into the customer's own sequences and messages, so the sending obligations sit with the customer's tooling. What is specific to gifting is the recipient eligibility problem.
Sending an item of value to a public official, a healthcare professional, a procurement officer or an employee of a regulated institution engages anti bribery and gift acceptance rules that vary by country and by sector, and nothing retrieved addresses screening, value thresholds, jurisdictional blocks or reporting. A platform whose whole function is transferring value to named individuals at other companies should speak to this. Ask what recipient screening and per recipient value caps the platform can enforce, and in which jurisdictions.
The instruments are published and a buyer cannot tell which of them governs. A privacy policy, an acceptable use policy, a marketplace vendor agreement and a subprocessor list are all linked from the footer, and the in product dashboard is described as allowing a customer to manage, update and delete personal information, recipient lists and order history.
Against that, the footer links two different sets of terms: one labelled Terms whose address marks it as a superseded version, and a separate link labelled Platform Usage Terms pointing elsewhere. The copyright line names a corporate entity under the former domain and a year already past. For a platform that collects home addresses from people who are not its customers, ambiguity about which agreement is in force is a substantive gap rather than a cosmetic one. Ask which terms document is the operative one and request it in writing.
There is no third party corpus here and nothing to license. The platform sells no contact database, no enrichment and no intent signal: recipients arrive from the customer's own record system or by import, and the delivery address is supplied by the recipient rather than bought from a broker, so the most sensitive field the platform holds originates with the person it describes.
The supply side is governed rather than assumed, with a published marketplace vendor agreement and a published route for suppliers to onboard, which means the goods catalogue has a stated contractual basis. It sits below the top band because the platform does accumulate a personal data set that did not exist before it: home and preferred delivery addresses volunteered by recipients across many senders.
Nothing published states how long those are held, or whether an address given to one sender is available to the next. Ask for the retention period on a recipient supplied address and whether it is reused across customers.
Exposure is low by architecture and the routes are sanctioned. The vendor operates its own marketplace, warehousing and fulfilment rather than borrowing access to anyone else's platform, so the primary flow is from vendor to customer. Integrations are named individually and run through each destination's own documented interface, covering two customer record platforms, a marketing automation system, two sales engagement tools, two scheduling products and a general automation connector.
The browser extension is distributed through the official extension store with a published listing rather than sideloaded, which is the sanctioned route for that delivery method and the point several vendors in adjacent categories fail. It stops below the top band for the reason most vendors in this position do: no conformance position of the vendor's own is stated anywhere, so the grade rests on architecture rather than on a documented commitment.
One stewardship question is open and it was opened by the acquisition. The acquirer's announcement of this purchase states that it will leverage a combined gifting database of past results to drive recommendation, and this platform's transaction history is part of what was bought.
Nothing on this product's surface states whether customer campaign data, recipient records or acceptance history feed that combined database, whether contribution is optional, or whether anything is aggregated across customers before it is used. A subprocessor list is published, which is the one stewardship artifact present, and it is hosted on the acquirer's domain rather than this one.
There is no model operating here today on the pages retrieved, so this is not a training disclosure failure but an unanswered question about where the data goes next. Ask for a written statement on whether your campaign and recipient data feeds the group's shared gifting database, and how to decline.
The recipient is a participant rather than a target, and what they are told stops at the gift. Nothing arrives unannounced: a notification reaches the person first, they are told a named company has sent them something, and they must actively accept and type their own delivery address before anything ships. There is no synthetic sender, no borrowed identity and no ambiguity about which company is behind it. Two things are not disclosed.
The accompanying note is composed by the sender or by the vendor's creative agency and nothing published states what, if anything, a recipient is told about how it was written. More materially, accepting requires handing a delivery address, often a home address, to a company the recipient has no relationship with, and the notification is not described as explaining who will hold that address, for how long, or how to have it removed. Ask what privacy notice the recipient sees at the moment they enter an address.
A named and checkable integration set rather than a claimed count. Destinations are listed individually and span two customer record platforms, a marketing automation system, two sales engagement tools, two meeting scheduling products and a general automation connector, with a dedicated integrations page behind them, so a buyer can verify their own stack before purchase rather than after.
The stated purpose is specific and matches the product: triggering a send from an event held in the record system and writing delivery and engagement status back against the contact. A browser extension is published on the official extension store with a live listing. What holds it below the top band is the developer surface: no public interface documentation, no webhook reference and no sandbox was reached on either pass, so anything beyond the prebuilt connectors is unevidenced. Ask whether a documented interface exists for sends and status callbacks, and whether access is gated.
Physical geography is documented and data geography is not. International shipping, global warehousing and overseas inventory storage are published as capabilities with partner logistics networks behind them, so a buyer can establish where goods move.
Where data sits is unaddressed on every page retrieved: no hosting region, no processing location, no residency election and no transfer mechanism for moving recipient personal data between jurisdictions, which matters more here than for a purely domestic product because the fulfilment network is explicitly international.
The one artifact bearing on this is the subprocessor list, and it is published on the acquirer's domain rather than this product's, so the processing chain has been consolidated into the parent while the product retains its own brand, domain and contract. Ask for the hosting region for recipient records and the transfer mechanism relied on for shipments outside the country of the sending account.
The security route is advertised and it does not go anywhere. A footer link labelled for security and compliance resolves to the homepage rather than to a security page, returning the homepage canonical address. The vendor's own pricing questions direct a buyer wanting security information to the legal section of the footer, and that section contains a privacy policy, an acceptable use policy, two sets of terms, a marketplace vendor agreement and a subprocessor list, and no security document of any kind.
So the signposting fails twice over, once at the link and once at the destination it recommends instead. No certification, examination report, audit period, auditor, penetration test, control enumeration or trust portal was located on either route. What is published, the subprocessor list, sits on the acquirer's domain, which shows the group can route a trust artifact across brands when it chooses to. Ask whether the parent's certifications cover this platform by scope, and for the certificate that names it.
The structure is published with unusual candour and every number is withheld. The vendor states the cost model plainly, an annual platform subscription plus the cost of the goods, tells a buyer the price varies by company size, user count and team count, and publishes one genuinely material commercial term that most of this category does not: charging only for gifts a recipient accepts, which removes the unclaimed gift from the bill and is the largest uncontrolled cost in gifting.
Against that, no figure appears anywhere. The pricing page shows a single plan, a feature list and a route to a sales conversation, and does not mention the free single user tier that the site navigation offers through a self serve signup link, so the one price a buyer could establish without talking to anyone is absent from the page named for prices. This is the more interesting form of the grade: the packages exist and are described, and the numbers alone are held back. Ask for the platform fee, the goods markup, and whether an unspent prepaid balance is refundable on termination.
In product control is stated and the way out is not. The account dashboard is described as letting a customer manage, update and delete personal information, recipient lists and order history, so a customer can act on their own records without raising a ticket, and deletion is named explicitly rather than implied. What is missing is everything about leaving.
No export format, no bulk extraction route, no interface for retrieving campaign and delivery history, and no statement of what happens to records after termination or how long a departing customer has to retrieve them. Delivery and engagement history is the asset a gifting programme accumulates and the reason attribution reporting has any value, so being unable to establish whether it can be taken out is the substantive gap. The ability to delete data is not the ability to keep it. Ask for a documented export of contacts, campaigns and delivery history, and the retention window after an account closes.
The hard problem in physical sending is designed out, and the soft one underneath it is unaddressed. Nothing ships to an address the vendor guessed: the recipient supplies their own at the moment of acceptance, which eliminates the stale address failure that wastes most of the budget in direct mail, and the vendor absorbs that risk commercially by charging only on acceptance.
Tracking is published by campaign, item, link and contact, with delivery status notifications and real time inventory visibility. The gap is that the whole model rests on one email arriving. If the acceptance notification lands in a spam folder the gift is never claimed, never shipped and never billed, and nothing published addresses the sending domain, authentication, inbox placement or a fallback route when the notification goes unopened. Ask for the acceptance rate and the notification open rate, and what happens to an unclaimed send after the reminder sequence ends.
Coverage is documented across buyer, use case and size rather than asserted. Four buying roles each have their own page setting out what the product does for them, covering marketing, sales, customer success and people operations, and that last one is a genuine distinction in this category because employee gifting and recruitment sit outside the revenue organisation entirely.
Three named solution lines address engagement, branded merchandise and account based programmes, with a separate enterprise track offering guided onboarding and a dedicated success manager. The commercial ladder spans a free single user tier through to a quoted enterprise arrangement, and international shipping and warehousing are published capabilities rather than a claim. Named customers span large technology firms and mid market companies. It stops below the top band because no country list, regional restriction or coverage limit is published anywhere despite international fulfilment being a headline capability.
Pricing
What this vendor charges, what it commits to in writing, and where the bill can move. Figures the vendor publishes itself are labeled Vendor Published. Figures labeled Estimated come from other sources and the vendor has not confirmed them.
- ›Postal does not tell you what it costs. There is a free version for one person that you can sign up for yourself, but the page about pricing does not even mention it, and the paid version has no price on it at all. To find out what you would pay, you have to book a call. The company does explain how the bill is put together: you pay a yearly fee to use the software, and then you pay separately for the actual gifts you send. The genuinely good part is that you are only charged for gifts people accept, so anything nobody claims costs you nothing.
How the price works
What you are charged for, and what makes the bill go up.
Annual platform subscription plus cost of goods, stated by the vendor as varying with company size, number of users and number of teams. Gift spend is funded through a prepaid balance topped up by card or invoice. The vendor states it charges only for gifts accepted by the recipient through the digital link, so unclaimed sends do not bill. A free single user tier exists and is not described on the pricing page.
What the contract says about your data
What the vendor commits to in writing once your data is in the product.
A subprocessor list is published, hosted on the acquirer's domain rather than this product's. A privacy policy, an acceptable use policy and a marketplace vendor agreement are published. Two separate terms documents are linked from the same footer, the primary one at an address marking it as a superseded version, so the operative agreement should be established in writing before signing.
Getting started
What it costs and what is included before the product is running.
Not published as a figure. Guided implementation, onboarding and a dedicated customer success manager are listed as features of the paid plan rather than priced separately. The vendor states typical onboarding completes within thirty to sixty days.
What to watch for
Where this pricing can surprise a buyer who has not read it closely.
Verified 4 September 2026 against the vendor's own pricing page and homepage. No figure is published at any level. The pricing page presents one plan, First Class, as a feature list with a route to a sales conversation and no number attached. It does not mention the free Economy tier, which is reachable only from a self serve signup link in the site navigation, so the single price a buyer could establish unaided is absent from the page named for pricing. The vendor does publish the cost structure in its questions section, stating an annual platform subscription fee plus the cost of the goods, and that price varies by company size, user count and team count.
It also publishes one material commercial term: charging only for gifts a recipient accepts through the digital link. Third party transaction data quoted by a procurement marketplace describes negotiated discounts and relative positioning against competitors but gives no list figure, so no dollar equivalent is recorded. Free tier described by third parties as a single user with domestic sending; that scope was not confirmed on the vendor's own pages, which do not describe the tier at all.