ActiveCampaign's native Salesforce sync is genuinely good, and it has specific limits — opportunity data moves one way, writes arrive in batches rather than instantly, custom object support is restricted. We design around what it actually does, then build the scoring and automation on top of it.
Trusted by 500+ organizations — RevOps, marketing operations and sales teams running ActiveCampaign alongside a CRM with Twopir Consulting.
Built For Sync You Can Actually Trust
None of these are ActiveCampaign faults. They are what happens when a capable platform is connected to a CRM without anyone deciding, in writing, what the connection is supposed to guarantee. Every one of them is a design decision that was never made.
Contacts get created in ActiveCampaign and never land in the CRM, or land late and inconsistently. Reps chase leads marketing already engaged last week, working from a picture that was accurate at some point and nobody can say when.
Scores get set once during onboarding and never revisited. Reps stop looking at them because the number no longer reflects what a prospect did, and a score everyone ignores is worse than no score — it looks like signal while carrying none.
Workflows get built during the initial setup and never touched again while the funnel changes underneath them. The segmentation now describes the business as it was two years ago, and everyone has quietly stopped believing the reports it feeds.
A contact becomes an opportunity, marketing loses visibility and sales loses the email and SMS history. Neither team can answer what an account has actually seen, which is exactly the question that decides the next conversation.
A new custom field appears on one side and does not map on the other. Data quietly stops syncing on that field alone, so nothing looks broken — until someone notices a segment is missing a third of its members, usually in a QBR.
It is not, and the documentation says so. Writes into Salesforce are batched, and inbound sync runs on a schedule you choose. That is a sensible design — it protects your API limits — but if an automation was built assuming instant reads, it will behave unpredictably and nobody will know why.
ActiveCampaign is a marketing automation platform that runs email, SMS, site tracking and behavioural triggers from a single automation engine, with a built-in scoring model and its own lightweight CRM. Its strength is the automation builder: one workflow can branch on what somebody opened, what page they visited, what they bought and what a CRM field says, without stitching three tools together to do it.
One boundary worth settling before anything is configured, because getting it wrong produces months of confusion: ActiveCampaign ships its own CRM, and you should not run two. If Salesforce is your system of record, then ActiveCampaign's deals pipeline is not — and the integration should be designed so marketing automation reads from and defers to Salesforce rather than maintaining a parallel truth beside it. Teams that leave both pipelines live end up reconciling them by hand, which is the exact problem the integration was bought to solve.
What ActiveCampaign ships is an automation and scoring engine. What decides whether it is trusted is the field-direction map, the scoring model and the suppression logic underneath — which is the work Twopir Consulting is engaged to do. Where a different platform is the better answer we say so: see Marketing Cloud and Account Engagement for the Salesforce-native options, Mailchimp for a lighter sibling, and marketing automation tools for how we compare them. ActiveCampaign documents the Salesforce integration itself in its help centre.
Most pages in this category describe the native connector as real-time and fully bi-directional. It is neither, and that is fine — as long as you design for it rather than discover it in month three.
| Capability | Native connector | iPaaS / middleware | Custom API build |
|---|---|---|---|
| Contacts, Leads & Person Accounts | Two-way, with field mapping configured on both sides. | Two-way, if you build and maintain both directions yourself. | Two-way, in whatever shape your process actually needs. |
| Opportunities | One way only — Salesforce into ActiveCampaign, to trigger and suppress. | Either direction, at the cost of another moving part to own. | Either direction, and you own every schema change on both sides. |
| When data lands | Writes into Salesforce are batched; inbound sync runs on a schedule you set. | Near-real-time is achievable, and consumes Salesforce API calls to get there. | As fast as you are willing to build and pay for. |
| Existing records | Historical data is not brought across by the sync; a one-time import handles it. | Backfill is a job you script once and then keep. | Backfill is part of the build, and so is re-running it. |
| Custom objects | Restricted. Standard objects are the supported path. | The usual reason teams add middleware in the first place. | Unlimited, and correspondingly the most expensive to carry. |
| Our usual advice | Right for most orgs — design around the edges above and it is reliable. | Right when custom objects or sub-batch latency genuinely drive revenue. | Right when the sync logic is proprietary and worth owning outright. |
Three genuinely different jobs, with three different buyers. On this product the middle one is by far the most common — ActiveCampaign is already running, and the CRM beside it tells a different story about the same person.
For teams standing ActiveCampaign up for the first time
A new build, or a migration from another automation platform. We design the contact and custom field model before importing anything, and we decide up front which system owns the pipeline so you never end up running two.
Delivered as configuration inside ActiveCampaign's own capabilities. Anything needing the API — a custom sync, a bespoke preference centre, a warehouse feed — is scoped separately, and we identify those at design time rather than discovering them mid-build.
For teams whose automation and CRM disagree with each other
Our most common entry point. Both systems work; the join between them does not. We audit both schemas, configure the connector against what it genuinely supports, and rebuild the scoring and automations on top of live CRM data.
The native connector will not carry custom objects or sub-batch latency. Where your requirements genuinely need either, we say so and scope the middleware or custom work honestly rather than promising the connector can stretch to it.
For teams whose requirements have outgrown the connector
Development against the ActiveCampaign and Salesforce APIs for what the connector cannot reach: custom object sync, sub-batch triggers where the revenue genuinely justifies them, and internal tooling that lets a non-marketer fire the right automation safely.
Every custom integration is something you own through every API change on both sides. We argue each one against the native alternative in writing first, because a bespoke sync nobody maintains fails more expensively and more quietly than the connector it replaced.
Every capability below is built around your CRM data model and the way your revenue team actually works — not around the connector's default settings, which assume nothing about either.
The map that decides which system wins, field by field, before a single record moves. Because the native sync runs in both directions and cannot be switched off in one, field ownership is not optional configuration — it is the thing that stops the two systems overwriting each other quietly.
A scoring model calibrated against your own engagement data rather than carried over from a template. Scores update on real behaviour — opens, clicks, page visits, form fills — and are reviewed on a schedule, because a model set once during onboarding is a model sales will have stopped reading by the second quarter.
Multi-step automations across email, SMS and site behaviour, branching on CRM field values as well as engagement. One workflow can react across channels, which is ActiveCampaign's genuine strength — and the reason a badly designed one causes damage across all of them at once.
Opportunity data flows one way — Salesforce into ActiveCampaign — and that one direction is exactly what you need to stop a nurture email reaching someone in final negotiation. Suppression is the single highest-value thing this integration buys, and it is the thing most setups never configure.
Email, SMS and site engagement surfaced on the record sales already works from, and campaign performance read against pipeline rather than against opens. A rep should be able to open an account and see what it has been sent without switching systems or asking marketing.
Since February 2024, Google and Yahoo require bulk senders — broadly, anyone sending around 5,000 messages a day to personal Gmail accounts — to authenticate with SPF and DKIM aligned under a DMARC policy, offer one-click unsubscribe, and keep spam complaints low. SMS carries its own consent rules alongside.
An automation built on a field that arrives ten minutes late behaves unpredictably, and the bug is invisible. Each integration below states its purpose, which direction data moves and how quickly it gets there.
The core sync, covering Leads, Contacts and Person Accounts. It runs in both directions and cannot be limited to one, which is precisely why field ownership has to be decided rather than assumed.
Salesforce ⇄ ActiveCampaign · both waysDeal stage flows into ActiveCampaign to trigger and — more usefully — to suppress. This leg is one-way by design, so ActiveCampaign should never be the place a pipeline change is recorded.
Salesforce → ActiveCampaign · one waySignup surfaces wired to create the ActiveCampaign contact and the CRM record in one pass, with owner assignment applied on entry rather than in a weekly tidy-up.
Form → contact and CRM · on submitSMS steps inside the same automation as email, on the same profile and the same consent record. One workflow reacting across channels is the platform's real advantage over a stack of point tools.
Automation → channel · same journeyShopify, WooCommerce and BigCommerce order and browse data in, so purchase-behaviour segmentation is real rather than inferred — and a repeat buyer never receives a first-purchase offer.
Store → ActiveCampaign · order eventsUnsubscribes and preference changes reflected on both sides. This is the one place where the batch window genuinely matters, so we design the opt-out path to be the fastest thing in the system.
Preference ⇄ both systems · prioritisedZapier, Workato and MuleSoft where custom objects or sub-batch latency are genuinely required. This is the honest route past the native connector's limits, and it is a system you then have to own.
Middleware ⇄ orchestrated, with retryEngagement exported so email, SMS, product usage and revenue sit in one model — and, where segmentation logic is genuinely complex, assembled in the warehouse and pushed back as a simple flag.
Engagement → warehouse · scheduledMost implementations run 4–6 weeks, depending on how many custom objects and existing automations need to be mapped and rebuilt. Plan entitlement and data cleanup are the two things that move that range, and both are settled in week one.
We map every object, field and existing automation across both platforms before touching anything, and we confirm your plan actually entitles you to the native connector. Both matter: the sync design has to account for what your teams really use, and a missing entitlement is a procurement cycle rather than a configuration change.
We install and configure the connector, set the inbound schedule against your API allowance, and build the field-by-field direction map that holds up when the schema changes. Existing records are backfilled deliberately, because the sync itself does not bring history across — and reconciled before anything goes live.
We rebuild the nurture, notification and scoring logic against your real funnel stages, then connect scoring outcomes to CRM routing and alerts. Deal-stage suppression is built here rather than added later, because the cost of getting it wrong is a nurture email landing in a live negotiation.
We test sync behaviour under real data changes — including the awkward ones, like a record edited on both sides inside one batch window — train marketing and sales on the shared record, and hand over documentation your admin can maintain without us.
Four patterns that come up repeatedly. These are illustrative scenarios, not client case studies — we would rather show you the shape of the work than attach numbers we cannot attribute. For published work in the same problem space on a different platform, see our marketing automation and Salesforce integration case study.
Trial signups engage with nurture emails and hit key product pages, and nobody in sales finds out until the weekly report. The work has ActiveCampaign update the CRM as those signals arrive, assign a follow-up task to the right rep, and start a sequence matched to what the person actually looked at.
Purchase behaviour has to drive both the email programme and the sales conversation. The work segments customers on real order data in ActiveCampaign while the CRM holds account-level revenue and renewal timing, with the two kept in step so a repeat buyer never receives a first-purchase offer.
Reps need to know the moment a lead's behaviour signals real intent rather than idle browsing. The work ties score thresholds to specific site and email actions and fires a CRM notification when one is crossed, so outreach is prioritised on signal instead of on who is nearest the top of the list.
Dozens of local teams need consistent follow-up without dozens of duplicated workflows. The work centralises first-touch nurture and lead routing while giving each location visibility of its own pipeline — corporate keeps control of the automation logic, local teams keep ownership of their deals.
If one of these reads like your quarter, the first call will be short and specific rather than a platform overview.
ActiveCampaign's own deals feature is live, Salesforce is live, and nobody decided which one is the system of record. Somebody reconciles them by hand, and that reconciliation has quietly become a job.
The model was configured at onboarding and never recalibrated. Rebuilding it against real closed-won outcomes, and agreeing the thresholds with the reps who act on them, is usually a two-week piece of work with a disproportionate effect.
It happens once, an account executive escalates it, and the fix gets logged as a marketing problem. It is a suppression architecture problem, opportunity data is already flowing in the right direction to solve it, and almost nobody has configured it.
Consistent follow-up across many local teams, with central control of the automation logic and local ownership of the deals. The architecture question is where the boundary sits, and it is worth answering before the tenth location goes live.
Anyone can install a connector. Whether it is still trusted a year later depends on decisions made before it was switched on — and those are the decisions we are hired for.
Most sync failures trace back to field mismatches nobody documented, and a bi-directional sync you cannot switch off in one direction will happily propagate them. We map both sides before configuring anything, because after go-live the same work is a data cleanup project.
Opportunity sync runs one way. Writes are batched. Custom object support is restricted. None of that is a flaw, and all of it changes how automations should be designed — so we put it in the scope document rather than letting you find it in month three.
Scoring models are calibrated against your own engagement data and reviewed against closed-won outcomes on a schedule. A template model carried over from another client is exactly how a score becomes a number sales ignores.
Marketing automation defers to live opportunity data, so campaigns never work against what sales is doing. That means deciding early that ActiveCampaign's own pipeline is not the pipeline — a decision that is cheap to make now and expensive to unwind later.
Field mappings, direction rules and automation logic are handed over in a form your admin can maintain. Twopir Consulting has delivered CRM and marketing operations for 12+ years, and the measure of a good handover is that the calls stop.
Bi-directional yes, real time no, and the distinction matters when you are designing automations. Contacts, Leads and Person Accounts sync in both directions. Opportunities do not — deal data flows one way, from Salesforce into ActiveCampaign, which is exactly the direction you need for triggering and suppression but means ActiveCampaign should never be where a pipeline change is recorded. On timing, ActiveCampaign writes to Salesforce in batches rather than instantly, documented as roughly every ten to eleven minutes or sooner once about a hundred contacts have changed, and the inbound sync from Salesforce runs on a schedule you choose ranging from hourly to once daily. That is a deliberate design rather than a limitation: Salesforce meters inbound API calls, and a chatty integration spends that allowance on behalf of every other system in your org. The practical rule we apply is that any automation which must react within seconds should be triggered by an ActiveCampaign-native event rather than by a Salesforce field it has to wait for.
Not an entry plan. The native Salesforce connector sits at the top of ActiveCampaign's marketing tiers — Starter, Plus, Professional and Enterprise — and at the time of writing it is an Enterprise capability. Vendors repackage regularly and comparison articles go stale, so confirm entitlement against your own contract rather than against any published summary, ours included. We check this in week one of every engagement, because it is the most common scoping surprise on this product: a team plans a six-week integration and discovers the connector needs a plan change first, which is a procurement cycle rather than a configuration change. If the tier is not justifiable for your volume, that is a legitimate reason to look at middleware or a different platform, and we will tell you so rather than quietly building around it.
Support for custom objects in the native connector is restricted — standard objects, meaning Leads, Contacts, Person Accounts and one-way Opportunity data, are the supported path, and custom fields on those standard objects map without difficulty. If your segmentation genuinely depends on a custom object, the honest answer is that the connector is not the mechanism for it, and the options are middleware such as Zapier, Workato or MuleSoft, or a direct API build. Before reaching for either, it is worth asking whether the value you need from that custom object can be flattened into a field on a standard object — a status, a tier, a calculated flag — because a single well-chosen field is enormously cheaper to run than a second integration platform. That question is part of the design conversation, not an afterthought once the connector has disappointed someone.
Not if the design accounts for them, which is why we audit existing flows, triggers and validation rules before configuring any sync. There are two specific risks worth naming. The first is a validation rule that rejects records the sync tries to write, which produces silent failures rather than errors anyone sees. The second is a Salesforce flow that fires on field update and now fires every time the sync touches that field — a loop that is harmless in staging and expensive in production. We map both before go-live and set field ownership so the two systems cannot fight over the same value. Since the sync runs in both directions and cannot be limited to one, that ownership decision is not optional.
Not automatically. Connecting the two systems starts the ongoing sync; it does not backfill what already exists on either side. Existing Salesforce leads and contacts need a deliberate one-time import, and that import is where the duplicates in your database become visible — which is uncomfortable but far better than discovering them a month later mixed in with new records. We treat the backfill as its own step with its own reconciliation: match rules agreed first, a dry run against a subset, then the full import, then a count check on both sides before automations are switched on. Skipping that sequence is the most common reason a new integration produces a contact database nobody trusts.
Most implementations run four to six weeks, depending on how many custom objects and existing automations need to be mapped and rebuilt. Two things move that range more than anything else. The first is plan entitlement: if the connector requires a tier you are not on, the timeline now includes a procurement conversation, which is why we check it in week one. The second is data condition — if the CRM already carries duplicates, the backfill has to wait on a cleanup, and that cleanup is a data project rather than a configuration task. We define scope, timeline and milestones before any configuration begins, and if the audit says the data has to be fixed first, that is what the plan will say rather than a shorter number that turns out to be wrong.
Pick one and mean it. ActiveCampaign includes a capable lightweight CRM, and for a small team without an existing Salesforce investment it is a reasonable place to run a pipeline. What does not work is running both — the moment two systems each hold deals, somebody is reconciling them by hand, and that is the problem the integration was bought to solve. If Salesforce is your system of record, the integration should be designed so ActiveCampaign reads deal stage and never writes it, which also matches how the connector works: opportunity data flows one way. If you are genuinely undecided, the deciding questions are usually whether sales forecasting, territory management and reporting need to live in the same place as the pipeline, and whether anything else in the business already depends on Salesforce data. We will give you a recommendation rather than a menu.
We audit what you already have — plan entitlement, field mapping and direction, sync scheduling, scoring model, suppression rules and engagement reporting — and hand back written findings with prioritised recommendations. If the honest answer is that the connector cannot do what you need, that is what it will say.
Response within 24 hours · We start with diagnosis, not a sales call · Contact the team
