Every Heed discovery call lately ends with the same question. The client lists out their tools, we describe what the custom app would do, they pause, and then they ask. "Wait, can you build that too?" The answer, almost always, is yes.
AI-fluent custom development has collapsed the cost of recreating SaaS features to the point where keeping seven separate subscriptions is now the expensive option. The features that used to require a multi-million dollar engineering team now take a focused build on top of a handful of foundation models, a connector layer, and a clean dashboard frame. We are not arguing that every SaaS should be replaced. We are arguing that the default has flipped, and the conversation now starts with "what do we keep" instead of "what do we add."
Below is the working list of feature categories we routinely fold into a single custom app, the named SaaS competitors clients are usually paying for, and what the recreation actually looks like.
The pattern is consistent. Generic-purpose SaaS that charges per seat for features you can describe in three sentences is exactly the category where custom now wins. Deep-domain industry vendors with 20 years of vertical knowledge are exactly the category where we connect, not rebuild.
1. Form Builders
Typeform, JotForm, Wufoo, Google Forms, Gravity Forms. Per-seat or per-submission pricing, branded with their logo, and limited to whatever fields and logic the vendor has shipped. We replace these with custom multi-step intake forms that live inside the dashboard, support conditional logic, push directly to your CRM record, and use AI to summarize free-text responses before a human ever opens the submission. Same UX as Typeform, your brand, your data, no per-submission cap.
2. Calendar and Scheduling
Calendly, Acuity, Cal.com, SavvyCal, Chili Piper. The core feature is a public booking page tied to your availability. We recreate that with branded gating, qualifying questions before the booking finalizes, round-robin routing across attorneys or project managers, and integration directly into the matter or project record so the meeting context is already there when the call starts. No separate vendor login, no $15 per seat per month.
3. E-Signature
DocuSign, Adobe Sign, Dropbox Sign, PandaDoc, HelloSign. Clients hate maintaining a separate vendor account just to sign a document, and the per-envelope pricing scales painfully. In-portal signing with a full audit log, IP capture, and timestamp meets the same legal standard. The client logs into your portal, signs, and the executed document lands in their case folder automatically. One login. One brand. No DocuSign bill.
4. Email Automation
Mailchimp, ConvertKit, ActiveCampaign, Klaviyo, Drip. The feature set is multi-stage drip tracks, behavior-triggered sends, segmentation, and template editing. We rebuild that as part of the dashboard, with one important upgrade: AI drafts the copy in your voice, pulled from past emails you have written, so the campaigns actually sound like you instead of like a marketing intern.
5. Landing Pages
Unbounce, Leadpages, Instapage, ClickFunnels. The pitch is "drag and drop your way to a high-converting page." The reality is that the highest-converting pages are the ones that match your existing brand exactly and pull from your real product data. We recreate the landing page builder as a section of the dashboard, with AI-drafted copy, A/B testing, and direct integration to the lead pipeline. No separate domain, no separate billing.
6. Survey and NPS
SurveyMonkey, Typeform NPS, Delighted, AskNicely. In-portal feedback collection with sentiment AI on every response, automatically routed to the right manager when a score drops below a threshold. Your customer never sees a third-party vendor logo on a feedback request, and your team gets a sentiment dashboard alongside every other piece of customer context.
7. Knowledge Base and Help Center
Zendesk Guide, HelpScout, Notion, Intercom Articles, Document360. The killer feature here is search across your own documentation. We rebuild that as an internal AI search across your existing M365 or Google Workspace files, no migration required. Your team asks a question, the AI cites the right document. No separate paywall, no migrating 800 articles into a new system.
8. Live Chat and Support
Intercom, Drift, Zendesk Chat, LiveChat, Tidio. We replace the generic chat widget with an Anthropic-powered assistant scoped to your business, your tone, your products, and your policies. It can read order history, draft refund decisions, escalate to a human when stakes are high, and log everything to your CRM record. The customer experience is closer to talking to a knowledgeable employee than a generic chatbot trained on the open web.
9. Project Management
Asana, Monday, ClickUp, Trello, Basecamp, Smartsheet. The features that matter are kanban boards, Gantt views, status reporting, and per-role visibility. We recreate those inside the dashboard with custom views per role, no per-seat tax, and project status that pulls directly from the systems where the work actually happens (Procore, Lawcus, NetSuite) instead of requiring duplicate manual entry.
10. CRM
Salesforce, HubSpot, Pipedrive, Zoho, Close, GoHighLevel. We have written extensively about CRM replacement elsewhere. The short version: most clients are paying for 20 percent of the feature set at 100 percent of the seat price, and AI-driven contact enrichment, deal scoring, and follow-up drafting close the gap on the rest. The structural engineering firm we work with was 85 percent migrated to Salesforce when they walked away and asked us to build something better.
11. AP, Billing, and Spend Management
Bill.com, Ramp, Brex, Stripe Billing, Tipalti. Custom AP intake reads incoming invoices via email, OCR plus AI extraction parses line items, three-way match runs against the PO and receipt, and the GL posting fires through the QuickBooks or NetSuite connector. The approval flow lives in the dashboard. The expense card layer (Ramp, Brex) is a place we often connect rather than rebuild, since the underlying card-issuance infrastructure is genuinely complex.
12. File Management and Digital Asset Management
Dropbox, Box, ShareFile, Hightail, WeTransfer. We use Cloudflare R2 as the storage layer, role-scope every file at the application level, and let AI tag files automatically based on content. One file system replaces three or four overlapping tools. Clients log in once and see only what they should see.
The legal firm we work with came in paying for DocuSign, Dropbox, a separate client portal SaaS, and a generic AI chatbot. The custom dashboard collapses all four into one secure portal under Cloudflare Zero Trust, with the bonus that high net worth client data never sits on a vendor server again.
13. Video and Loom-Style Recording
Loom, Vidyard, Bonjoro. In-app video record, automatic AI transcript, send-to-record routing so a project manager can request a 90-second video update from a field assessor with one click. The video lives in the project record, not in a separate vendor account that you lose access to when the employee leaves.
14. Mass Email and Outbound Outreach
Apollo, Outreach, Lemlist, SalesLoft. Outbound sequencing tied directly to your CRM data, with AI-drafted personalization that actually references the prospect's website and recent news. The email-sending infrastructure (deliverability, warm-up, suppression lists) is something we typically connect to via an SMTP provider rather than rebuild, but the campaign engine, AI personalization, and reply tracking all live in the dashboard.
15. Analytics and Dashboards
Looker, Tableau, Domo, Sigma, Mode, Power BI. The output is real-time business KPIs custom-rendered for each role. The owner sees revenue and margin. The PM sees project status and budget burn. The field assessor sees jobs assigned and reports due. We render those views directly inside the same dashboard as everything else, sourced from your live operational data, with no separate BI tool to license and train people on.
Where SaaS Still Wins
To be honest about the tradeoffs: there are categories where standardization is genuinely valuable and we connect rather than rebuild every time.
Stripe, for payment processing. The compliance, fraud detection, and global-rails infrastructure is decades of engineering you cannot replicate, and Stripe's marginal cost is reasonable. We connect to Stripe and surface payment data inside your dashboard. We do not try to recreate a payment processor.
Shopify, for ecommerce checkout. The conversion-optimized checkout flow, the global tax and shipping integrations, and the payment infrastructure are also decades of work. We connect to Shopify and surface orders, inventory, and customer data inside the dashboard. We do not try to rebuild a checkout.
AWS, Azure, and Cloudflare, for hosting. We deploy on top of these. We are not trying to build a cloud provider.
Twilio, for SMS and voice infrastructure. Per-message pricing is reasonable, and the underlying carrier relationships are genuinely hard to replicate.
Industry-specific deep-domain vendors. Procore for construction. Famous Software for fresh produce distribution. Lawcus for legal practice management. AppFolio for property management. NetSuite or SAP for serious manufacturing ERP. athenahealth or Epic for clinical workflows. These tools have 15 to 30 years of vertical knowledge baked in. We connect, surface their data inside the dashboard, and add an AI layer on top. We do not try to recreate them.
How We Decide What to Recreate vs. Integrate
The mental model we use on every discovery call has five inputs.
First, how standardized is the workflow. If the workflow is identical across every business that uses the SaaS (payment processing, checkout flows, EDI transactions), keep the SaaS. If the workflow is customized to your business, recreate.
Second, what is the per-seat cost relative to recreation cost. If you are paying $30 to $80 per seat per month for features that can be specified in a paragraph, the math usually says recreate inside an annual build budget that pays back in under 12 months.
Third, does the SaaS lock in your data. If exporting your data is hard or impossible, that is a structural risk and a reason to migrate to a custom system you control.
Fourth, does AI add disproportionate value if recreated. Email automation, customer support, knowledge search, and form intake all become dramatically better with AI in the loop. Stripe with AI is not meaningfully better than Stripe.
Fifth, is there a deep-domain industry vendor we cannot match. Procore knows construction better than we ever will. Famous Software knows produce distribution. Connect, do not compete.
The result of running those five questions across a typical 12-tool stack is almost always the same: keep two or three deep-domain or commodity-rails tools, recreate the other nine inside one dashboard, and let AI add a layer on top that none of the original SaaS could ever deliver alone.
The Limit Is Imagination
We have a phrase we use internally that has started leaking into client conversations: "limited only by imagination." It is not marketing. It is a description of where the cost curve actually sits in 2026. The features clients used to outsource to seven vendors at $30 per seat per month are the same features we now bundle into one custom app for less than the annual cost of two of those subscriptions.
If you want to see what your stack would look like as one custom dashboard, the right next step is a discovery conversation where we walk through what you currently pay for and what we would actually recreate. We will tell you honestly which tools are worth keeping and which ones are worth rebuilding. The math usually makes the decision easy.
More on the architecture is on the Apps & Dashboards page and the Operations Platform overview. Two case studies are worth reading: the structural engineering firm that walked away from Salesforce, and the law firm dashboard for a high net worth practice.