# Loops vs SendGrid Source: https://wraps.dev/versus/loops-vs-sendgrid A tool built for one kind of company against a platform built for everyone. The narrow one is faster and will run out of room. The broad one will do whatever you ask and make you configure it first. One of these was designed around a specific customer — a software company with subscribers, onboarding sequences and a product update to send — and everything in it assumes that shape. The other was designed to be configurable enough for anybody, and carries the accumulated compromises of twelve years of keeping many kinds of customer satisfied. Choosing is mostly a question of whether you are the customer the narrow tool was drawn for, and whether you expect to still be that customer in three years. ## At a glance |   | Loops | SendGrid | | --- | --- | --- | | What it is | Product + marketing email for SaaS, priced per contact | Twilio's email platform — API, SMTP, and marketing | | Published pricing | Free at 1,000 contacts and 4,000 sends/mo. Paid plans price on subscribed contacts with unlimited sends — $49/mo in the 1K–5K band, $99/mo in the 5K–10K band, rising from there. | Essentials $19.95 at 50K and $34.95 at 100K. Pro $89.95 at 100K, $249 at 300K, $499 at 700K. Dedicated IP is Pro and above; extra IPs $30/mo. Marketing Campaigns is priced as its own plan, not an add-on. | | Best for | SaaS teams who want lifecycle and transactional in one tool and never want to count sends again. | Enterprises wanting one procurement relationship, SMTP everywhere, and marketing in the same console. | | Watch out for | You pay for contacts, not sends, so a large dormant free list stays on the bill. It is a product rather than a primitive, with less low-level control than an email API. | The free tier died on 27 May 2025, replaced by a 60-day trial at 100/day. Support latency and abrupt account suspensions dominate its public reviews. | Prices are each vendor's published list prices, verified August 2026. Check the vendor's own page before you sign anything. ## Where they actually differ ### Who the product was drawn around Loops A software business with subscribers: onboarding sequences, feature announcements, trial nudges, a monthly update. Because the assumed customer is that specific, a great deal of configuration simply does not exist — there is one obvious way to do most things and it is usually the right one. A team outside that shape will feel the absences quickly. SendGrid Everyone, which is a strategy rather than an accident. An enterprise with subusers, a store, an agency, an internal tool sending one alert a day: all of them are supported, and the cost of supporting all of them is a surface where nothing is opinionated and most things need configuring before they work. ### What the bill is charged against Loops Subscribed contacts, with sending unlimited. You stop counting messages entirely, which changes behaviour — nobody hesitates over an extra send because it does not cost anything. The trade is that a large dormant free list stays on the invoice forever whether you mail it or not, so list pruning becomes a billing decision. SendGrid Sending volume, on a plan ladder, with the marketing product priced as its own plan rather than an add-on. A quiet list costs nothing until you mail it. The structure rewards a business with many contacts and infrequent campaigns and punishes one that sends often to a small, engaged audience. ### Product email and campaigns in one place Loops Transactional sends and lifecycle campaigns live in the same tool against the same contact record, which is the consolidation that makes Loops attractive to a small team. One vendor, one list, one view of a person, and an engineer triggering a password reset is using the same system the marketer schedules an announcement in. SendGrid Both exist under one account and they are separate products with separate plans, separate interfaces and separate mental models. That is more capability and less consolidation: the procurement relationship is single and almost nothing else about the experience is. ### Who opens it on a Monday morning Loops Anybody on the team. The interface is modern, the concepts are few, and a founder or a marketer can build and send something without a walkthrough. That accessibility is the main reason small teams choose it, and it is a real operational saving when there is no dedicated email person. SendGrid Somebody who has learned it. SendGrid's marketing interface is capable and carries a decade of accumulated options, and getting a campaign out involves knowing which of several similar-looking things you want. On a larger team that is fine, because the person who knows it is the person who does it. ### How far a segment can go Loops Audience filters on contact properties and on events you send in, which is enough for the segments a SaaS team actually uses: on a trial, signed up this month, has not completed setup, on the paid plan. It stops well short of multi-condition behavioural logic, and the moment you want a rule with three nested branches you have found the ceiling. SendGrid Segments on contact fields and engagement, with the depth limited by what you put into the fields. It is more configurable than Loops and no more behavioural, because the underlying model is still a contact with attributes rather than a stream of events. Both stop in the same place, arriving there from different directions. ### The developer surface Loops A small, current API with SDKs, aimed at identifying contacts and sending events from your application so the tool can react. The surface is deliberately compact and that is appropriate for the product's scope. It is not a general-purpose sending platform and does not pretend to be. SendGrid A large mature API, SMTP relay, webhooks with rich event metadata, tagging and custom arguments for slicing analysis later, and a decade of third-party integrations built against it by other people. At volume this is the more capable half of the pair by some distance, and the price is an interface somebody had to learn. ### Systems that only speak SMTP Loops Loops is an API-and-interface product. If something in your estate can only be pointed at an SMTP host — a CMS, a monitoring agent, an appliance — that traffic is not going here and needs somewhere else to go. SendGrid SMTP is a first-class path and has been forever, which is why so much software has a SendGrid option in a dropdown. A great deal of infrastructure was configured against it years ago by people who have since left, and that inertia is a genuine reason accounts stay, independent of whether the product is the best choice today. ### What happens when an account is questioned Loops A smaller vendor with a smaller support queue, which usually means faster and more personal answers, and a shorter operating record to judge it by. There is no widespread pattern of complaint about arbitrary account action, which counts for something in a category where sudden suspension is the dominant fear. SendGrid Support latency and abrupt suspension dominate SendGrid's public reviews, consistently enough to plan around. Large platforms act first when a sending pattern looks unusual against a shared pool, and a campaign is exactly that pattern. Keep a fallback configured and know in advance how you would reach a human. ### Where each one stops being enough Loops You outgrow Loops when your messaging needs to branch on behaviour in ways a filter cannot express, when you need per-tenant sending isolation, or when a legacy system needs an SMTP host. Those are real ceilings and they arrive at a predictable size, which is an argument for choosing it deliberately as a stage rather than as a forever decision. SendGrid You outgrow SendGrid's marketing half when you need genuine behavioural journeys, at which point you add a lifecycle platform and keep SendGrid for sending underneath it. That is the arrangement a great many companies land on, and it is worth recognising as the likely destination when you are pricing the marketing plan today. ## Pros and cons, side by side ### Loops Strengths - Lifecycle campaigns and product email run against one contact record, so the engineer firing a password reset and the marketer scheduling an announcement are looking at the same row rather than reconciling two. - Sending stops being a number anybody watches. Teams that quit metering messages run more experiments and more nudges, because nobody has to justify the marginal send to anybody. - Whole categories of configuration are simply absent, and for a company of five that is a saving counted in weeks rather than in clicks. There is usually one way to do a thing and it is usually right. - A founder or a marketer can build and send something the first time they open it, which is a genuine operational saving when nobody on the payroll is an email person. Trade-offs - The invoice tracks people you keep rather than mail you send, so a dormant free-plan list is a standing charge and pruning becomes a budgeting conversation rather than a hygiene one. - There is no relay at all, so anything in the estate that can only be handed a hostname and a password needs a permanent home somewhere else, which is a second vendor rather than a workaround. - Each transactional email has to be authored and published in the product before any code can reference it, which reverses the usual order of work and puts a marketing tool on the critical path of a backend deploy. - Audience filters read contact properties and events and stop where a three-branch behavioural rule begins. The ceiling arrives at a size you can predict, and it does arrive. ### SendGrid Strengths - A relay has been a first-class path here for over a decade, which is why so much software nobody on your team wrote already has this vendor waiting in a dropdown. - Subusers hand a platform real per-tenant credentials, standing and reporting, and the narrower product has nothing resembling it available at any price. - The event stream carries categories and custom arguments, so a finance or analytics team rebuilds its own view of the traffic instead of asking the vendor for a report it does not offer. - Twilio's balance sheet is itself a procurement answer, and for a certain kind of buyer that is a hard requirement rather than a stated preference. Trade-offs - The permanent free tier ended in May 2025 and became a sixty-day trial, so every recommendation written before that date was formed under economics that no longer exist anywhere. - Marketing Campaigns is a separate plan with its own contact database, so the one-vendor story arrives as two subscriptions metered against two different nouns that never reconcile. - Support latency and abrupt suspension are the two themes that dominate its public reviews, consistently enough to be a planning input rather than a rumour to discount. - Getting a campaign out means knowing which of several similar-looking objects you want, which is fine on a team with an email specialist and expensive on a team without one. ## Pick in thirty seconds ### Choose Loops if - You are a small software team and the emails you send are onboarding, announcements and product updates. - You want product email and lifecycle campaigns against one contact record rather than two products. - Not counting sends is worth more to you than not paying for dormant contacts. - Nobody on the team is going to learn a platform, so the tool has to be obvious on first open. ### Choose SendGrid if - Something in your estate needs an SMTP host and rewriting it is not on anybody's roadmap. - You send on behalf of customers and need per-tenant credentials, suppression and reporting. - Your list is large and quiet, so paying per contact for people you rarely mail would be the worse deal. - Procurement prefers a vendor inside a large public company, and that preference is a constraint rather than an opinion. ## What actually moves when you switch In roughly the order you would do it. The expensive parts of an email migration are never the send call. 1. Do the arithmetic in both directions first, because these two bill against different nouns entirely. One counts people on the list whether or not you ever mail them; the other counts messages, and then counts marketing contacts again on a second plan. The same company lands cheaper on either side depending only on the ratio between how many addresses it keeps and how often it writes to them. 2. Count the relay senders and give them a home that sits outside this decision. The narrower product has no SMTP endpoint, so the CMS, the monitoring agent and the invoicing system are not migrating anywhere — they are being rehomed to a third vendor, with their own credentials and their own subdomain, on a timeline nobody put in the original estimate. 3. Reconcile two incompatible shapes of consent before importing a single address. One side holds an account-wide unsubscribe alongside group suppressions that only bite when the send call names the group; the other holds a subscribed flag on the contact plus membership of each mailing list. Flattening the first into the second loses what somebody actually opted out of, and the cautious default of treating every suppression as global quietly stops receipts as well. 4. Author and publish every transactional email on the destination before the code that calls it exists. Loops binds a transactional send to a published template and a declared set of variables, so the deploy order reverses: content first, integration second. A team used to shipping the send call and filling the copy in later meets this in the sprint where it blocks them. 5. Rebuild the campaigns rather than exporting them. Automations, single sends and audience filters are three distinct objects on one side and a flatter, smaller set on the other, so every sequence gets re-decided by a person. That is a calendar item with a named owner, not a script somebody runs overnight. 6. Prune on engagement, not on age, before importing toward the per-contact meter. Every address that has ignored you for a year becomes a line on the invoice the moment it arrives, and stepping a plan back down after the fact is considerably harder than never stepping up. 7. Retire the link-tracking and authentication records on a tail, not on a date. Both sides rewrite tracked links through a subdomain they control, so pulling the record kills every link inside mail already sitting in inboxes. Leave the old records published until your longest-lived email has genuinely gone cold. 8. Decide who answers did that send afterwards, because the two tools answer it for different people. One hands a marketer a campaign report; the other hands an engineer an event stream with categories attached. Whichever way you travel, somebody's current method of answering that question stops working on cutover morning and nobody has told them. ## Questions to ask before you sign Put these to a sales rep, or find the answer in the documentation before anyone quotes you a number. 1. Is our list large and quiet or small and mailed constantly, and has anybody divided the one number by the other yet? 2. What in our estate can only be given an SMTP hostname and a password, and who owns each of those systems today? 3. Do we need a marketer who can publish without a deploy, or an engineer who can send without waiting on a marketer? 4. When somebody unsubscribes from the product newsletter, what should still be allowed to reach them, and which system enforces that? 5. Are we the same shape of company in three years, or are we buying a stage — and if it is a stage, what does leaving it cost? 6. Does anybody here send on behalf of our own customers, now or on the roadmap, and what happens if one of them buys a list? 7. If the account were suspended on a Friday afternoon, who would we contact, and is there a second path already configured and tested? ## A third shape of answer Notice what this comparison is not about: neither side is being chosen for how well it puts mail in an inbox, because both are adequate at that and the decision is really about the layer above. If the sending underneath is the part you would rather own — because the bill scales with it, because the retention window is short, or because you want the same infrastructure to survive changing your marketing tool twice — the third shape is to run SES in your own AWS account and point whichever lifecycle tool you pick at it. Wraps is one way to operate that with tooling on top. It does nothing for the campaign-building half of this page, and it needs an AWS account and SES production access, an approval on AWS's schedule that can be refused. Wraps publishes this page. We are not one of the two vendors being compared here, and this note is on it because the tension above is one owning the AWS account actually addresses — not because it is on every page. ## Common questions ### Is unlimited sending really unlimited? Within a fair-use posture, yes, and the honest framing is that the constraint moved rather than disappeared. You are no longer metered per message; you are metered on how many people are on the list. That genuinely changes behaviour for the better — teams that stop counting sends run more experiments and split more tests — and it means the wrong-shaped customer, one with an enormous list they rarely mail, pays a lot for the privilege of not counting. ### What happened to SendGrid's free tier? It was retired in May 2025 and replaced by a sixty-day trial. A very large number of small projects, side businesses and internal tools had been running on that tier for years, and all of them had to make a purchasing decision inside the same window. If you are reading a comparison written before that date, its recommendations were formed under different economics and should be re-checked rather than trusted. ### Can I run marketing on Loops and transactional on SendGrid? Yes, and for some teams it is the sensible split — the legacy systems keep their SMTP host, the product emails go somewhere pleasant to work in. Use separate subdomains so the reputations are independent, and decide explicitly which system owns the authoritative consent record. Two systems each keeping their own view of who unsubscribed is how somebody stops receiving order confirmations because they opted out of a newsletter. ### How do I know if my segmentation needs are too complex for Loops? Write down the three most valuable messages you want to send and the exact condition each depends on. If the conditions are properties and simple events — is on a trial, has not finished onboarding, upgraded last week — a filter-based tool covers them. If any of them contains the word and followed by a time window and a negation, such as did X but not Y within fourteen days, you are describing a behavioural journey and you should evaluate tools built for that instead of hoping to grow into one. ### What does migrating between these two actually involve? Contacts and their consent status export from both, and that is the straightforward part. Campaigns and automations have to be rebuilt by hand because nothing crosses the paradigm. Templates likewise. The piece most often underestimated is the application integration: every place your code identifies a contact or fires an event has to be rewritten against the other API, and the forgotten one is usually in a background job rather than in the main request path. Run both in parallel for a cycle with the new system receiving events but not sending, so you can confirm the data is arriving correctly before anybody's inbox depends on it. ### Does Loops give me an SMTP host I can point things at? No, and that is a scope decision rather than a gap they intend to close. Loops is an API and an interface, built on the assumption that the thing sending is your product. Anything in your estate whose only email configuration is a hostname, a port and a password — a WordPress install, a backup appliance, a monitoring agent, an ERP — cannot be pointed at it at all. That traffic needs a relay, which means a second vendor and a second subdomain regardless of what you decide about the product email. Work out how many of those systems you have before you price anything, because the answer is almost always more than the first count found. ### What happens to my unsubscribe groups when I move to Loops? They flatten, and the flattening is lossy in a direction worth understanding. SendGrid lets a message name a suppression group, so somebody can be opted out of your product newsletter while remaining perfectly reachable for a billing notice. Loops models a contact as subscribed or not, plus membership of each mailing list. Mapping several groups onto mailing lists works, but only if you decide first which of your old groups represented a marketing preference and which represented a genuine do-not-contact. Getting that backwards either re-mails somebody who asked you to stop or silently drops a receipt they are entitled to. ### How does each bill behave in a month when we send nothing? Loops charges the same as any other month, because the meter is the number of subscribed contacts you are storing. SendGrid's transactional plan charges its plan fee and no volume beyond it, while Marketing Contacts keeps charging for the audience regardless. So a quiet quarter costs roughly full price on the per-contact side and noticeably less on the per-send side — which is exactly the inverse of a heavy campaign month. If your sending is seasonal, model a full year rather than a representative month, because a representative month does not exist for you. ### Can either of these send on behalf of our own customers? SendGrid can, properly, and that capability is the single hardest thing to replace in this pair. Subusers give each tenant its own credentials, its own sending reputation, its own suppression list and its own reporting, so one customer importing a bad list does not degrade delivery for the rest. Loops has verified domains on a flat account and nothing equivalent, so multi-tenant isolation becomes either separate Loops accounts you administer by hand or an acceptance that your tenants share standing. If you are a platform rather than a single business, this dimension usually settles the comparison before anything about template editing gets discussed.