Pipedrive is built around one idea and does it well: a deal sits in a stage, and it carries a next activity. What it will not do is fill itself in. We connect the phone, the calendar, the web forms and the billing, so the board reflects what actually happened rather than what somebody remembered to type.
Wired to the rest of your stack
Pipedrive works because the model is simple. One deal, in one pipeline, in one stage, with a value and a next activity attached to it. Rotting thresholds turn a deal red when nobody has touched it for long enough, and that pressure is the reason sales teams actually keep the board current. All of it assumes a person is typing.
The calls do not type themselves in. Neither does the enquiry that arrived at nine in the evening, the quote that was accepted, or the invoice that finally cleared. None of those become an Activity, and none of them move a deal a stage. So the board goes red on deals somebody has been chasing all week, and stays green on ones nobody has looked at since Tuesday, and the forecast is built on top of that.
Pipedrive’s own automations close part of the gap and then stop. They are genuinely good at the simple rule: a stage change, an owner assignment, a follow-up activity created the moment a deal moves. What they will not do is loop over a set of records, branch into several real outcomes, or wait on something happening in another system. The allowance is counted per company as well, not per seat, so it runs out sooner than a growing team expects. We build the part that sits past that line and write the result back into Pipedrive, so nothing on the board depends on anyone remembering.
An inbound call is matched to a Person by phone number and written up as a completed Activity on their open deal, with the transcript filed as a Note. Pipedrive stores phone numbers as a labelled list with one marked primary, so we match on every value rather than only the first and the same customer does not end up as two records.
A Lead in Pipedrive cannot sit in a pipeline until it is converted into a Deal, which is why unqualified enquiries so often go quiet in the inbox. We set the routing and the conversion rule at intake, so raw enquiries land as Leads, qualified ones land as Deals in the right stage, and neither waits on somebody noticing.
Every custom field is addressed by a 40-character key rather than its label, dropdown answers come back as numeric option IDs, and a monetary field carries its own separate currency key beside it. We map against those keys, so renaming a field in Settings changes what your team sees and nothing else.
Anything needing a loop, more branches than your plan’s if/else allowance, or a wait on another system gets built in n8n and writes back through the API. Your Pipedrive automation count stays free for the stage-change rules it handles well and your team can still edit on their own.
Pipedrive gives your whole company one daily API allowance that scales with your plan tier and your seat count, and search calls are the most expensive requests in it. We cache what does not change, batch the writes, listen on webhooks instead of polling, and use the v2 endpoints, which cost fewer tokens than the ones they replace.
Coming from a spreadsheet or a flat CRM means deciding what becomes an Organization, what becomes a Person, what becomes a Deal, and what was only ever a Note. We map that first, de-duplicate on email and phone, then run a dry pass you can check field by field before anything is written into the live account.
Use Pipedrive’s own automations wherever they fit, because they are visible in the account and your team can edit them without calling us. They handle a single trigger and a straight run of actions well: a stage change, an owner assignment, a follow-up activity created the moment a deal moves. They run out when you need to loop over a set of records, branch into several genuinely different outcomes, or wait on something happening in another system, and the number of if/else conditions inside one automation is capped by your plan. There is also a loop protection layer that stops runs without writing anything to the history, so an automation that quietly died looks identical to one that had nothing to do. We keep the obvious rules inside Pipedrive and build the rest outside it, which keeps the count available and the complicated logic somewhere you can actually read it.
No, and it catches people out at renewal. The automation allowance is counted per company, not per seat, so a team of twelve on a given plan gets exactly the same number as a team of two. Custom fields work the same way, one pool for the whole account shared across deals, people, organisations and products. What seats do increase is your daily API allowance, which is calculated from your plan tier multiplied by your seat count. So hiring gives you more integration headroom and no extra automations, which is usually the opposite of what a growing team assumes it is buying.
Everything talking to Pipedrive starts getting a 429 back, across the whole company, until the budget resets at Pipedrive’s server midnight rather than yours. It is a daily pot rather than a per-minute throttle, so one wasteful sync can spend it before lunch and take every other integration down with it for the afternoon. Search requests are both the most expensive calls in the system and the most tightly throttled, which matters because caller lookup and de-duplication lean on search harder than anything else. We design around that: cache what does not change, batch the writes, prefer webhooks over polling, and use the v2 endpoints rather than the older ones. Pipedrive sells budget top-ups from a certain tier upward, but paying for headroom to cover a badly built integration is the expensive way to fix it.
Often it is not, and that is worth saying before you spend money making it fit. Pipedrive is a sales CRM and the deal is the centre of everything: one pipeline, one stage, one linked Person and one linked Organization, unless you start adding person-type custom fields to hold the rest. There is no native concept of a recurring contract, a monthly run rate or a renewal cycle, post-sale delivery lives in the Projects add-on which is charged separately on the lower tiers, and a support ticket has to be bent into a Deal or a Project because no object exists for it. If most of your revenue is repeat work, renewals and service against the same customers, you will be arguing with the data model every month and we would rather tell you now than after the invoice. If your revenue comes from named opportunities that get won or lost, Pipedrive is one of the better places to run them, and connecting it beats replacing it.
This is the failure mode almost nobody knows about, and it is silent. Pipedrive counts your active leads together with your open, won and lost deals against a single company-wide cap. Go past it and a deal created by an automation, an import or the API is not created at all: it is parked on a deals waitlist instead. Leads created automatically are worse, because they are discarded outright with nothing to recover them from. Your integration logs a clean run, the record simply is not there, and that is why we alert on the count rather than waiting for somebody to notice a missing enquiry.
Usually one of three things, and none of them announce themselves. A personal API token belongs to a user, so when that person leaves and their account is deactivated, every integration built on their token dies with them, which is why we authenticate through an OAuth app or a dedicated account instead. Pipedrive also deletes a webhook outright if it goes three consecutive days without a single successful delivery, so an endpoint that was down over a long weekend does not simply resume, it has gone. Third, Pipedrive is retiring its older v1 endpoints, so anything built against them, including some Zapier steps and older n8n nodes, has to be moved across. We monitor for all three, because the point of an integration is that you stop watching it.
Pick the tier for the allowances rather than the feature list, because the allowances are what an integration actually collides with. Automations do not exist at all on the entry plan, the number of if/else conditions permitted inside one automation rises with the tier, custom fields are capped per company, and your daily API budget is multiplied by the tier as well as by your seat count. Add-ons such as LeadBooster, Campaigns, Projects and Caller are charged on top of the plan and mostly per seat, so the headline price is rarely the number you end up paying each month. We will look at what you want built and tell you the lowest tier that carries it. If the honest answer is that your plan will not, we would rather say so at the start than build something that hits a ceiling in month two.
Send us your pipeline structure and the list of tools sitting around it. We will tell you what can be written straight into Pipedrive, what has to live outside it, and what your current plan will actually allow.
Book a Free Demo