IPS Is Not a Report Card, But a “Practical Guide”

When many automotive companies first receive their IPS report, their initial reaction is to flip to the last page to check the total score——and then either celebrate or panic.

But this is a classic “report card mindset” pitfall.

The value of IPS lies not in telling you “what score you got,” but in telling you “why AI gave you this score” and “how this score is currently affecting your potential customers”.

 

Understanding the Causal Chain——The “Five-Tier Transmission” from AI Recommendations to Sales Growth

Before dissecting the report, you must first understand a fundamental question: Why can IPS predict sales?

In traditional marketing theory, the consumer decision path is summarized by the AIDMA model:

> Attention → Interest → Desire → Memory → Action.

 

The implicit premise of this model is that consumers actively "discover" brands in the ocean of information, gradually build preferences, and ultimately make a purchase.

But in the AI era, this premise has collapsed. Users no longer "discover" brands; instead, they "receive" AI's recommended lists. Users don’t need to remember—AI remembers for them; users don’t need to painstakingly compare dozens of brands—AI has already sorted them out. The "Memory" stage in the AIDMA model is almost completely skipped in the AI era.

Based on the shift in consumer habits, we have upgraded the CISCAS full chain adapted to the AI environment:

> Cognition → Interest → Search → Comparison → Action → Share

Car companies cannot directly buy the recommendation slots of large models, but based on IPS monitoring results, they can fully acquire the brand labels and evaluation tendencies output by AI externally, and use this as a basis to adjust online reputation materials, thereby guiding large models to provide more favorable brand recommendations.

 

What is the first step when getting an IPS report?

Once you understand the causal chain, the "reading method" of the report becomes clear. Working backwards from the logic of "how AI affects users", first locate which link in the causal chain the problem lies in, then strike precisely.

 

• Check the total score: Locate the brand's AI recommendation "rank"

What to look at: IPS composite score (0-100 points).

How to look at it: Place it in a coordinate system to see the performance of key indicators across multiple dimensions and brands. The score level can indicate how much AI values you.

 

• Deconstruct VS/RS -- diagnose which link the "cause" is in

What to look at: Sub-scores of VS (Visibility) and RS (Priority). Different causes require completely different remedies.

These two dimensions form a 2×2 diagnostic matrix:.

	VS High	VS Low
RS High	✅ Ideal State: AI knows you and is willing to recommend you	⚠️ Problem: AI occasionally mentions you but has weak recommendation willingness—"knows you but doesn't care about you"
RS Low	⚠️ Problem: AI frequently mentions you but doesn't prioritize recommending you—"knows you but doesn't recommend you"	❌ Most Dangerous: AI hardly mentions you—"doesn't know you at all"

Practical Advice:  If VS is low, prioritize solving "AI visibility"—optimize your brand's exposure in sources frequently crawled by AI; if VS is high but RS is low, the issue lies in "content credibility"—need to restructure the authority and structured level of brand information.

 

• Look at the dimension radar chart: Among the 9 major dimensions, which dimensions are reasons for AI's recommendation to you, and which are weaknesses

What to look at: 9 major dimensions' itemized scores.

Core Dimension	Interpretation Direction
Safety	AI whether it considers your car "safe enough"
Appearance	AI how it describes your looks and interior
Driving Experience	AI evaluation of driving experience such as handling, chassis, power, etc.
Range/Energy Consumption	AI how it mentions you in the context of range
Intelligence	AI whether it classifies you as "technology-leading"
Value Retention	AI thinks you are "worth preserving"
Cost Performance	AI thinks you are "worth the price"
After-sales Service	AI's evaluation of your after-sales network
Safety	AI's evaluation of your safety

 

9 major diagnostic dimensions correspond to the factors most concerned by consumers when making car purchase decisions. Through the test of dimension-specific recommendation questions, you can clearly see whether the brand is recommended by AI because of "good quality" or because of "high cost performance", thereby guiding resource allocation.

 

• Competitive Benchmarking——Your "Actual Rank" on the AI Shelf

What to look at: Competitive benchmarking: the rank of direct competitors and the gap between you and them.

Many times, the brand's problem is not "what it did wrong," but "what competitors did right." If a competitor scores much higher than you in the "smart driving" dimension, and your "smart driving" parameters are actually not bad, the problem may be that the AI has not read detailed enough technical information from you. 

The core question answered at this step: Did I lose due to my own shortcomings, or due to my competitor's strengths?

 

 

• Look at audience differences—When encountering target users, did the AI say the right thing for you?

What to look at: Differences in AI recommendations among users of different age groups, to determine whether the target audience is effectively covered by AI.

If the brand's target users are "tech enthusiasts aged 25-35," but IPS shows that when AI recommends to this group, the main reason is "cost-effectiveness"—this is an AI persona mismatch and must be corrected immediately.

If the brand's target audience is “luxury car buyers aged 40-50 looking to upgrade”, but IPS shows that AI almost never recommends the brand to this demographic——this is a target audience coverage gap that needs to be addressed through content sources and EEAT development.

The core question this step answers: AI not only matters in terms of “what it says”, but “who it says it to” matters just as much——are the reasons your target customers hear from AI aligned with the brand positioning you want to convey?

 

• From analysis to action——developing a P0/P1/P2/P3 priority strategy

After completing the above five steps, consolidate the diagnostic results into a priority matrix that directly guides GEO execution: content production, source building, official website optimization, and multi-model monitoring, forming a complete closed loop of “diagnosis→execution→re-monitoring”. GEO is not about optimizing a website's ranking in search engines, but about optimizing the probability of a brand appearing in AI answers. Its core logic is: AI favors authoritative, specific, and verifiable information.

 

The IPS report is not the finish line, but the starting point for “brand AI health management”

The market competition in the current automotive industry has long moved beyond traditional dimensions such as hardware specifications, offline channels, and conventional advertising. As consumer decision-making interfaces fully shift to AI Q&A scenarios, a brand's cognitive image and recommendation ranking in the AI world have become key factors influencing customer acquisition and sales. The IPS report serves as a professional health check for brands within the AI ecosystem.

Leveraging the CISCAS full-chain user decision-making to precisely identify communication pain points, it conducts layer-by-layer breakdown analysis from five major dimensions: overall score, VS/RS core indicators, nine-dimension radar chart, competitive benchmarking, and audience segmentation. It scientifically prioritizes optimization from P0 to P3, implements iterative improvements in an orderly manner, and ultimately transforms AI-driven positive recommendation advantages into precise customer acquisition, actual orders, and long-term user reputation.