Building a Custom CRM for a Renewables Company: What We Learned

Key Takeaways

• Off-the-shelf CRMs like HubSpot and Salesforce are built for generic sales pipelines, not the multi-stage, survey-heavy workflows of a renewables installer.
• Teratherm Energy needed a system that handled quoting, job management, and customer data in one place - not three separate tools stitched together with manual effort.
• Custom-built does not mean slow: Aucta AI's delivery from operational audit to working system runs in 2 to 4 weeks.
• The biggest source of operational leakage in renewables businesses is the gap between an enquiry arriving and a qualified quote leaving (not bad salespeople).
• A custom CRM can be built to reflect how your business actually works, rather than how a software vendor imagined a generic business might work.

Why Off-the-Shelf CRMs Fail Renewables Businesses

The honest answer is that most CRMs were not designed for the way a renewables installer actually operates. Salesforce is an enterprise sales tool. HubSpot is a marketing-led platform. Pipedrive is built around a linear deal pipeline. Each of those tools does what it says on the tin - but a renewables business like Teratherm Energy does not work in a straight line from lead to closed deal.

Think about what a typical job actually involves. A homeowner enquires about a heat pump or solar installation. That enquiry needs to be qualified: is the property suitable? Is the customer on the right tariff? Are there ECO4 scheme eligibility considerations, or is this a fully self-funded job? Then there is a site survey to schedule, often with a third-party assessor. Then a quote that depends on the survey findings. Then MCS certification requirements to track. Then install scheduling, sign-off, and aftercare. That is not a pipeline with five stages. That is a branching, multi-department workflow with data flowing between surveyors, office staff, and installation teams.

When you try to map that into HubSpot, you immediately start bending the tool. You add custom fields until the interface becomes cluttered. You build workarounds using Zapier to push data between HubSpot and a separate job management tool like Jobber or Tradify. You start keeping a spreadsheet on the side because neither tool gives you the view you need. Before long, your CRM is no longer a source of truth - it is one of three places you have to check before you can answer a customer's question about their job status.

This is exactly where Teratherm Energy found themselves. The data was spread across too many places. Quoting was slow because whoever was responsible for producing the quote had to pull information from multiple sources before they could start. Jobs were progressing, but nobody had a clear, single view of where each job stood at any moment. And the weekend admin - chasing updates, reconciling records, preparing reports - was mounting up. The business was growing, but the operational infrastructure was not keeping pace.

What makes this particularly sharp in the renewables sector is the regulatory dimension. MCS certification, Ofgem requirements, ECO4 eligibility assessments, and grid connection paperwork are not optional extras. They are part of every job. A generic CRM has no concept of these. You can add fields to track them, but the system does not know that a job cannot move to install stage until MCS documentation is complete. It does not flag when a compliance step has been missed. It just holds the data passively and relies on a human to notice the gap. In a business doing ten, twenty, or thirty installs a month, that reliance on human vigilance is where things start to fall through.

What We Actually Built for Teratherm Energy

The starting point was not software. It was an operational audit - working through exactly how Teratherm Energy's jobs moved from first enquiry to completed install, where data was created, who touched it, and where the friction was. That process surfaces the specific gaps that a custom system needs to solve. It also stops you building the wrong thing, which is the most expensive mistake in any software project.

What emerged was a clear picture of the core problem: Teratherm needed a single environment where a job could live from lead to aftercare, with the right information visible to the right person at each stage. Not a sales CRM bolted onto a separate job management tool. One coherent system, built around the actual workflow.

The custom CRM and ERP we built for them handles the full job lifecycle. When an enquiry arrives, it enters the system and gets triaged automatically. Relevant qualification data - property type, heating system, ownership status, potential ECO4 eligibility - gets captured at the point of first contact, not chased up later. That information then travels with the job through every subsequent stage, so the surveyor sees what the office knows, and the quoting process starts from a complete picture rather than a blank sheet.

Quoting was a particular area of focus. In the old setup, producing a quote required manually gathering survey notes, calculating equipment costs, checking availability, and formatting a document that looked professional enough to send to a customer. That took time, and time between survey and quote is where jobs go cold. The system we built connects the survey data directly to the quoting engine. When a surveyor completes their assessment, the relevant fields populate automatically, and the quote is generated from a template that already knows the job details. The person responsible for approvals reviews it, makes any adjustments, and sends it. The manual assembly step disappears.

Job management flows from the same data. Once a quote is accepted, the job status updates and the install scheduling process begins without anyone having to re-enter information that already exists in the system. The MCS documentation requirements are built into the job workflow as mandatory checkpoints - the system does not allow a job to progress past certain stages until the compliance steps are recorded as complete. That is not a human remembering to check a list. That is the process enforcing itself.

We also built and designed Teratherm's website as part of the same project, with AEO and GEO architecture built in from the start, and deployed an email enquiry handling agent that ensures incoming leads are captured, categorised, and actioned without anyone having to manually triage an inbox. The CRM, the website, and the email agent are not separate tools - they are one connected system where data flows in one direction and nothing gets lost in transit.

The custom software and dashboard layer means Teratherm's team has a single view of the business at any given moment. Pipeline by stage, jobs by install date, quotes outstanding, compliance checkpoints pending - all visible without running a report or opening a spreadsheet. That operational visibility is not a luxury. When you are managing multiple installs across a region, knowing where every job stands without asking five different people is what keeps a business functional at scale.

What a project like this demonstrates is that the question is rarely "which CRM should we buy?" The more useful question is "what does our business actually need to do, and does any existing product do that without us bending ourselves around it?" For businesses with genuinely complex, multi-stage workflows - and every serious renewables company or construction firm has them - the honest answer is usually no. Off-the-shelf tools are built to be flexible enough for the average business. If your business is not average, that flexibility has a ceiling, and you hit it faster than you expect.

The operational leakage that builds up around that ceiling is real. Slow quotes, missed follow-ups, compliance steps that rely on one person's memory, weekend admin to reconcile records that should already be reconciled - these are not minor inconveniences. They are the difference between a business that scales and one that grows headcount just to keep up with its own paperwork. A custom system built around workflow automation principles does not just reduce admin. It removes entire categories of manual work from the equation.

When Does Custom Actually Make Sense, and When Does It Not?

This is the question worth answering honestly, because a custom CRM is not the right answer for every business. Knowing the difference saves you a significant amount of time and money.

If your business is genuinely early-stage, doing fewer than five or six jobs a month, and your workflow is relatively linear, an off-the-shelf tool will probably serve you well enough for now. Jobber works well for straightforward trades scheduling. Pipedrive is fine if your sales process really is just lead, call, quote, close. HubSpot's free tier can handle basic contact management without much friction. The point is not that these tools are bad. It is that they have a specific design assumption baked in, and if your operation matches that assumption, you are fine.

The inflection point comes when your workflow has enough complexity, enough branching, enough compliance requirements, or enough data moving between enough people that you are spending meaningful time managing the tools rather than managing the business. For Teratherm Energy, that point had already arrived. The volume of jobs, the regulatory requirements of the renewables sector, and the number of people touching each job had pushed them past what any single off-the-shelf product could handle without significant manual workaround.

A useful test: count how many places a new enquiry gets recorded before it becomes a live job. If the answer is more than one, you have a data fragmentation problem. Count how many times information gets manually re-entered from one system into another during a job's lifecycle. If the answer is more than twice, you are paying for that duplication in staff time every single week. Count how many times a job has stalled or a follow-up has been missed because it fell between two systems or two people. If you can name specific examples without thinking hard, the cost of that leakage is almost certainly higher than the cost of fixing it properly.

The contra-indication is equally important to state clearly. A custom system requires that you actually know your process. If your workflow is still changing every few months, if you are still figuring out how you handle surveys or quoting or scheduling, building a custom system now will lock in a process that you will want to change six months later. In that situation, use a configurable off-the-shelf tool, settle your process, and build custom once you know what you are actually building for. Custom is a precision tool. It works best when you have something precise to build around.

The Specific Workflows That a Custom Build Unlocks

There are operational capabilities that simply cannot be replicated in an off-the-shelf CRM without building something that looks custom anyway. The most valuable are those where data from one stage of a job needs to automatically shape what happens at the next stage, without a human acting as the connector.

Take satellite property analysis as a concrete example. For a solar installer, knowing the roof orientation, pitch, and approximate usable area before a surveyor visits changes the economics of a sales call entirely. A system that pulls that data automatically at the point of enquiry, appends it to the lead record, and uses it to pre-populate sections of the quote template is not a feature you will find in HubSpot. It requires a custom data pipeline connecting external property data sources to your internal job record. Aucta AI's satellite property analysis service is built specifically to sit inside this kind of workflow, so surveyors arrive with context rather than starting from scratch on every visit.

Similarly, consider what happens when a quote goes out and receives no response. In most CRMs, the follow-up is a task assigned to a person. That person has thirty other tasks. The follow-up happens late, or not at all. A custom system can automate that follow-up sequence entirely, sending a timed email at day three, a WhatsApp message at day five, and flagging the job for a personal call at day seven, all without human intervention. That is not a feature of any standard CRM. It is a custom automated nurture workflow triggered by quote status, and it directly addresses one of the most consistent sources of lost revenue in any trades or renewables business: quotes that went out and were never chased properly.

The quoting process itself is another area where custom builds create a measurable operational advantage. When survey data automatically populates a quote template, when equipment costs pull from a live price list, and when the quote document is generated and sent for digital approval without anyone having to format a PDF, the time between survey and quote drops significantly. That compression matters because the customer's attention and buying intent are highest immediately after a survey. Every day that passes between survey and quote is a day the competitor's quote might arrive first. Speed is not just an efficiency metric here. It is a commercial one.

The estimating and quoting capability we built for Teratherm connects those stages directly. Survey fields feed the quote engine. The quote engine generates a professional document. Approval triggers job creation. No copy-paste. No reformatting. No waiting for someone to get around to it. The process moves at the speed of data, not the speed of whoever's inbox is busiest.

What the Build Process Actually Looks Like

One of the consistent concerns businesses raise before a custom project is how long it takes and how disruptive it is. The assumption is that custom software means a long discovery phase, months of development, and a painful cutover from old systems. That has not been our experience, and it was not the case for Teratherm.

Our process starts with an operational audit. That means mapping the actual workflow: how enquiries arrive, how they are qualified, who touches them at each stage, where data lives, and where the friction is. That audit is not a theoretical exercise. It produces a specific list of what the system needs to do, in the order it needs to do it. It also surfaces the things a business thinks are problems but are not, and the things they have accepted as normal that are actually costing them significantly.

From there, we build against live requirements rather than a speculative brief. The system is developed iteratively, which means the business can see and test real functionality within the first two weeks. By week four, the core system is operational. The goal is never to build everything at once. It is to get the highest-impact workflows running first, so the business starts recovering time and reducing leakage as quickly as possible, then add layers from there.

What this means practically is that the disruption is lower than expected, because the team is not being asked to learn a new system in a vacuum. They are being shown a system built around the workflow they already understand, with the manual steps removed and the data they need in front of them. Adoption is faster when the software reflects how people actually work rather than requiring them to adapt to how the software thinks they should work.

If you want to understand where your own operation is losing time and money before committing to anything, the AI automation checklist is a good starting point. It covers the specific operational signals that indicate whether a custom build is the right next step, or whether there is a simpler fix that would move the needle first. If you already know you need to build something properly, the right move is to talk to us directly.