Artificial Intelligence

How much it costs to implement artificial intelligence at a dealership

AI for dealerships does not have a single price because there is no single product. Some solutions charge per user, others per voice minute, others per conversation and others by stock volume. Comparing proposals without understanding that structure leads to buying what looks cheapest and ends up being the most expensive. This article explains what the real cost is made up of and how to calculate the return before committing.

Carlos Horno

Carlos Horno

CEO & Co-founder

12 min

Index

Share

Index

  1. The five cost items nobody mentions on the first call

  2. Real price benchmarks in Spain in 2026

  3. Why a low rate can end up expensive

  4. How to calculate total cost per use case

  5. Three calculation scenarios: low, expected and high

  6. Worked example: voice receptionist with eight appointments a month

  7. When to pay per user, per usage or per platform

  8. Costs that are usually left out of the sales price table

  9. How to request a comparable quote

  10. Dealcar and the cost of AI built into the DMS

  11. Frequently asked questions


Dealcar: stock, leads, portals and website, all connected. Request a demo

The five cost items nobody mentions on the first call

The price shown on the pricing page of any AI solution is only part of the real cost. The five items that make up the full cost of implementing AI at a dealership are the licence, usage, implementation, integration and operation.

The licence is access to the platform or agent: the published monthly or annual fee. It is the only cost you can see before talking to anyone.

Usage is the variable cost that depends on actual use: call minutes in voice solutions, number of messages in WhatsApp solutions, number of images processed in photography tools, or tokens in text-generation solutions. When usage exceeds the plan you signed up for, there are overages billed separately.

Implementation covers the initial setup of the system: designing conversation flows, writing voice or chat scripts, migrating data from the previous system and training the team. It can be a one-off cost of several hundred euros or be included in the first months' fees, depending on the provider.

Integration is the cost of connecting the AI solution with existing systems: the DMS or management system, the CRM, the appointment calendar, the website and telephony. If the integration requires custom development, it can be a significant cost that does not appear in the price table.

Operation is the recurring cost of keeping the system running well: reviewing conversation quality, updating scripts when the sales process changes, technical support, managing phone numbers and WhatsApp templates (which have their own cost in Meta Business), and the team's time spent supervising the system.

Real price benchmarks in Spain in 2026

The prices below are those officially published by each provider in September 2026, excluding VAT where stated. They are not a substitute for a commercial offer and may have changed.

Emblematic — Voice solution for handling calls. Entry plan: 99 euros a month for 1,000 voice minutes. Team plan: 299 euros a month for 3,000 minutes. Custom Enterprise for higher volumes.

voz.ag — Voice agents and conversational AI. Plan at 499 euros a month for 1,500 minutes. Plan at 999 euros a month for 4,000 minutes. Published overage of 0.08 euros per additional minute.

Walcu — CRM with automation features for dealerships. Base plan: 99 euros per user per month. Sales Pack: 157 euros per user. VoIP Pack: 197 euros per user. Advanced automation may be quoted separately depending on the project.

Dealcar — DMS with CRM, stock, publishing and built-in AI features. Plan up to 30 cars: 99.90 euros a month. Plan up to 70 cars: 129.90 euros a month. Plan up to 100 cars: 159.90 euros a month. AI buying and sales agent features are in development.

Read the 25 AI use cases for dealerships to identify which use cases deliver the most return.

Why a low rate can end up expensive

A solution with a low monthly fee can have high variable costs that make it more expensive than a solution with a higher fee but with usage included. The most common example is voice: if the plan includes 1,000 minutes and the dealership receives 1,400 minutes of calls a month, the 400 minutes of overage are billed separately. With an overage of 0.08 euros per minute, that is 32 euros extra. At 0.20 euros per minute, it is 80 euros extra.

On the other hand, a broader or more expensive solution may be unnecessary if the use case is simple. If the only goal is to recover missed calls outside business hours, a solution at 499 euros a month with 1,500 minutes may be overkill for a dealership that receives 200 calls a month.

The comparison should be made on the annual cost of the complete use case (licence plus estimated usage plus amortised implementation plus operation), not on the monthly rate of the cheapest plan.

How to calculate total cost per use case

The formula that matters is: net monthly cost = licence + estimated usage + monthly operation. If you want to calculate the return, subtract the value of the time or margin recovered: net monthly cost minus (time recovered multiplied by the team's hourly cost) minus (incremental sales multiplied by the average margin).

To apply that formula, you need to estimate actual usage before signing up. For a voice solution, usage in minutes depends on the number of calls the dealership receives outside opening hours and the average length of each conversation. If the dealership receives 80 out-of-hours calls a month with an average length of 4 minutes, the estimated usage is 320 minutes a month.

For a description and photo generation solution, usage depends on the number of vehicles listed per month and the number of photos per vehicle. If the dealership lists 20 cars a month with 15 photos each, usage is 20 descriptions and 300 processed images.

See also which AI tools are worth it for a used car dealer.

Three calculation scenarios: low, expected and high

To avoid surprises on the invoice, the cost should be calculated in three scenarios before signing up.

The low scenario uses 60% of the forecast volume. It helps you understand whether there is a guaranteed minimum or a lock-in period that forces you to pay even if usage is lower. If the plan has a monthly minimum of 99 euros regardless of consumption, that is the cost floor.

The expected scenario uses the estimated real average volume. It is the scenario on which the return is calculated and against which proposals from different providers are compared.

The high scenario adds 30% to the expected volume to simulate a peak month (August, September with the back-to-school period, December). The cost of overage in that scenario reveals whether you need to change plan in high season or whether the overage is affordable.

For each scenario you need to calculate the full monthly cost (licence plus overage plus operation) and check that it makes economic sense even in the high scenario.

See also AI integrated into DMS vs standalone AI tools.

Worked example: voice receptionist with eight appointments a month

A dealership signs up for an AI voice solution to handle out-of-hours calls. The monthly cost breakdown is as follows: 499 euros for the 1,500-minute plan licence, 150 euros for integration with the CRM and calendar amortised over 12 months, 80 euros of additional consumption for excess minutes during peaks, and 5 hours of system review and supervision by someone on the team at 25 euros an hour, which comes to 125 euros. Total monthly cost: 854 euros.

The system generates 12 incremental appointments a month (buyers who would have been lost with no out-of-hours response). Of those 12, 8 turn up. Of the 8 visits, 2 end in a sale with an average margin of 1,800 euros. Monthly incremental revenue: 3,600 euros. Return on cost: more than 4 times the investment.

If none of the 8 visits turns into a sale in the first month, there is no return yet. That is why the success metric should not be calls answered or appointments generated, but the incremental sales that can be attributed to the system. That attribution requires flagging in the CRM which leads came through that channel and how many ended up buying.

Read also how to structure a dealership's sales funnel.

When to pay per user, per usage or per platform

The per-user pricing model makes sense when each person on the team actively works in the tool and the number of users grows with the business. A CRM priced per user is reasonable if each salesperson has their own diary and their own lead pipeline.

Usage-based pricing (per minute, per image, per conversation) makes sense when the value of the tool is directly proportional to consumption. A voice tool that charges per minute is reasonable because the cost grows when the dealership grows and drops when business is quiet.

A platform or stock-volume fee makes sense when the whole operation shares the same database and the number of users is not the variable that changes most. A DMS that charges by the number of cars in stock grows with the inventory, which is the metric most directly linked to the size of the business.

The variable to analyse is which of those metrics (users, consumption or stock) will grow most when the business scales. If the plan becomes very expensive as you grow, the pricing structure may not be the right one for that particular business.

Costs that are usually left out of the sales price table

There is a list of costs that rarely appear in the first pricing presentation but that are part of the real cost of any AI solution.

The phone number, or porting the existing number, has a set-up cost and a monthly cost. Forwarding calls from the main number to the AI system may also have a cost depending on the carrier.

WhatsApp Business templates have a cost on Meta that varies by country and by type of conversation (business-initiated or user-initiated). In Spain, marketing conversations initiated by the business have a per-conversation cost that is added to the cost of the tool.

Find out how the WhatsApp integration with the Dealcar CRM works.

Building custom integrations with systems that have no native integration may require bespoke development. That cost can be one-off or recurring if the integration needs maintenance.

Cleaning up CRM data before migration or go-live has a time cost that is rarely budgeted.

Exporting data and migrating at the end of the contract can also have a cost if the provider does not offer that as a standard feature.


With Dealcar, what used to take you a day now takes one click. Request a demo

How to request a comparable quote

To be able to compare proposals from different providers for the same use case, you need to ask for each offer to include the same items at the same level of detail.

The information you need to give the provider to get a real quote is: monthly volume per channel (number of calls, WhatsApp messages, vehicles added per month), the number of users, sites and phone numbers, the exact actions the system must carry out in the DMS or CRM (create a lead, assign a salesperson, log the conversation), the languages and service hours, and the specific integration needed with your current systems.

The items you should ask to be included in the quote: set-up and training cost, monthly support cost, cost of consumption overages, minimum contract length and exit terms, and whether there is a trial period and on what terms.

Dealcar and the cost of AI built into the DMS

When the CRM, stock, website and AI features are in the same system, part of the integration cost disappears because the data is already connected. You don't have to pay for development so that the customer service tool knows the available stock, because it already has it in the same system.

That doesn't mean everything is included in the base fee: you need to ask which AI features are included in each plan and which are add-on services. Transparency on that distinction is part of what you should assess when comparing providers.

You can check the up-to-date plans and prices at dealcar.io/software-concesionario or request a demo where the starting point is a real car from your stock and a real lead from your current process.

Frequently asked questions

How much does it cost on average to implement AI in a small dealership selling 20 cars a month?

For a dealership of that size with basic needs (out-of-hours lead handling, description generation and photo enhancement), the total monthly cost including licence, consumption and operation is usually between 150 and 400 euros a month, depending on the provider and the actual usage volume. Below that, solutions are usually generic tools with no stock integration.

Is it worth asking for a free trial before committing?

Yes, but with enough scope to measure real results. A 14-day trial with low volume can show that the technology works, but not how much it will cost in production with real consumption or whether the system generates incremental sales. Ask for at least 30 days with the real call or lead volume from the previous month, and agree in advance which data will be compared at the end of the period.

How do I know whether the return justifies the cost?

By defining before you start which financial result will be attributed to the system and over what period it will be measured. Intermediate metrics (calls answered, appointments generated) are not the end result: the metric that justifies the cost is the incremental sales that can be attributed to the channel with reasonable certainty.

More from the blog

Keep reading

More ideas to sell more cars, manage better and depend less on portals.