The brief
A used-car dealership’s competition isn’t just the lot next door. It’s every similar vehicle that appears online, changes price, gets a better description, or disappears from the market.
That market moves quickly. A car priced well in the morning can look less competitive by the afternoon. A vehicle that seemed hard to position can suddenly look attractive once comparable inventory sells. Without current context, pricing and merchandising decisions tend to rest on incomplete information.
Giving a US used-car dealership that context is what Drishti was built to do. It continuously monitors competitor inventory, compares similar vehicles, studies dealership-level patterns, and turns changing market conditions into practical recommendations for the vehicles already on the dealership’s lot.
Why Drishti was built
A simple idea kicked things off: a dealership shouldn’t have to make inventory decisions without knowing what the surrounding market is doing.
Useful signals stream constantly from competitor websites — which vehicles are available, how similar vehicles are priced, how much mileage they have, which features and specifications they highlight, how complete their listings are, which dealers seem to be carrying particular types of inventory, and how the market changes over time.
Turning that scattered, changing information into something a dealership could actually use was the real challenge. The question Drishti was designed to answer sits behind all that data:
Given the market today, how should we think about each vehicle on our lot?
From websites to decisions
Publicly available inventory information gets continuously collected from competitor websites and turned into structured market intelligence.
The complete workflow is:
Monitor → Collect → Clean → Match → Compare → Interpret → Recommend
This goes beyond a one-time scraping exercise. Repeated observation of the market lets changes in availability, pricing, listings, and competitor behaviour become part of the analysis. What results is a living view of the competitive landscape rather than a report that’s outdated the moment it’s generated.
Monitoring a market that keeps changing
Constant motion describes competitor inventory well. New vehicles get listed. Prices get updated. Descriptions get rewritten. Sold vehicles disappear. Websites change their layouts and may load important information dynamically.
Operating in that environment is what Drishti was built for. Browser automation and specialised scraping tools navigate dynamic websites, collect relevant listing details, and feed the results into the processing pipeline — designed for repeated collection, not just a single successful extraction.
Comparing vehicles at the right level
Make and model alone don’t make a useful comparison. Two vehicles sharing the same model name may differ significantly in year, mileage, trim, specifications, condition, equipment, or listing quality, so Drishti compares vehicles across multiple attributes to identify more meaningful competitors.
For a vehicle on the dealership’s lot, the system can help answer:
- How many genuinely comparable vehicles are available?
- Is this vehicle priced above or below its competitive set?
- How does its mileage compare?
- Which specifications distinguish it from nearby alternatives?
- Are competing listings presenting information that our listing does not?
- Has the vehicle’s market position changed since the previous analysis?
A far more useful picture emerges than a simple average price would offer.
Understanding the vehicle’s position
Collecting competitor prices for their own sake was never the point. Understanding where each vehicle stood was.
A vehicle may be competitively priced but have a weak listing. Another may have a strong description and attractive specifications but sit priced above comparable inventory. A third might become more competitive simply because similar vehicles have left the market. All these signals come together in Drishti so the dealership can review a vehicle’s broader position rather than judging it on price alone.
Turning market signals into recommendations
Someone still has to decide what to do with raw market data. That’s where an intelligence layer comes in, turning observations into possible actions — surfacing signals such as a vehicle priced noticeably above or below comparable listings, comparable inventory becoming scarce, listings missing useful specifications, a competitor entering the market with a very similar vehicle, or a vehicle’s competitive position changing since the last collection.
That’s the shift from what is happening? to what deserves attention, and why? The final decision stays with the dealership; Drishti just cuts down the guesswork needed to make it.
Looking beyond individual listings
Comparing one vehicle with another is only part of competitive intelligence. Broader profiles of competing dealerships also get built by analysing their inventory over time — the types of vehicles they typically stock, their average pricing position, their inventory mix, the brands and segments they focus on, how their available inventory changes, and where their stock overlaps with the dealership’s own lot.
Two connected views come out of this: vehicle-level intelligence (how does this car compare?) and dealer-level intelligence (how does this competitor operate?).
A market that stays current
Only one moment gets described by a static analysis. Tracking change was the design goal for Drishti instead. Each collection creates another view of the market, letting the system observe vehicles appearing, disappearing, changing price, or being repositioned.
Over time, these observations build into a more useful competitive history — context that matters when deciding whether a vehicle is genuinely underperforming, temporarily exposed to more competition, or becoming better positioned as alternatives leave the market.
Engineering behind the pipeline
Extracting one listing from one website was never the hard part of this project. Building a repeatable system was — one that could navigate changing and dynamically rendered websites, collect information consistently, handle differences between competitor sites, normalise vehicle data into comparable fields, match vehicles that were similar but not identical, track changes across repeated collection cycles, and feed structured results into analysis and recommendations.
Browser automation, scraping reliability, data processing, entity matching, infrastructure, and intelligence logic all had to work together to make that happen.
The complete system
From public web data to dealership action, Drishti connects the entire path:
Competitor websites → Automated collection → Data normalisation → Vehicle matching → Market comparison → Dealer profiling → Recommendations
Outcome
A continuous competitive-intelligence capability for a US used-car dealership is what Drishti delivered — combining automated competitor monitoring, vehicle-level comparison, pricing analysis, listing-quality review, dealership profiling, historical market awareness, and recommendation workflows in one system.
Occasional manual competitor checks gave way to a more current, structured view of the market.
Project highlights
- Built a competitive-intelligence platform for a US used-car dealership.
- Delivered a live monitoring foundation within 10 weeks.
- Continuously tracked competitor inventory and listing changes.
- Compared dealership vehicles against relevant competitor vehicles.
- Analysed pricing, mileage, specifications, and listing details.
- Added vehicle-level market-position analysis.
- Identified listing information that may be missing or incomplete.
- Generated recommendations for pricing and inventory review.
- Built profiles of competing dealerships.
- Supported historical awareness of market changes.
- Used Playwright and specialised scraping tools for dynamic websites.
- Developed a production-oriented collection and analysis pipeline.
- Combined web data, entity matching, AI/ML analysis, and business recommendations.