The core finding: replying is not quoting
The store we measured replies fast, and even so the price takes a while (20-month medians):
| Stage of the conversation | Measured time |
|---|---|
| First reply in the chat | median 2 min (56,197 first replies) |
| First PRICE in the chat | median 16 min (31,352 requests with a price) |
| Requested parts unquoted in the first 24 hours | 1 in 5 (20%; Jul-Aug 2026) |
What the pattern shows
The pattern is clear: answering is immediate and quoting is not, because between the "let me check" and the price sits the slow part of the job: identifying the part, cross-referencing the number, checking stock and putting the answer together. The chat shows up in seconds; the search is still done by hand.
And it is not a bad streak: across 20 straight months (January 2025 to August 2026, 31,352 requests with a price) the monthly median for the first price moved between 13.2 and 18.5 minutes, with no trend. While request volume multiplied by 2.7, the time never improved: this is not a bad month, it is how a human counter works.
A canned reply confirms it
💡 The detail that settles it: the operation has a canned reply that reads "Your quote may take a few more minutes to be ready, given the nature of the part." When the apology for the delay already has its own macro, the delay is not an incident: it is the process.
Want to see how Victoria quotes a part like this? Message us on WhatsApp and we show you live.
Message us on WhatsApp →Half-finished quotes go cold
On multi-part orders, the customer has to chase the forgotten lines: "I am still missing the oxygen sensor, please." A partial quote nobody finishes is a sale going cold.
How the counter customer asks for a quote today
- The photo is the language: 1 in 3 requests arrives with a photo (29.6%: 15,520 of 57,681 requests over 20 months).
- The part number barely exists on this channel: at least 95 out of every 100 requests arrive without one, and probably 97. The rest is plain language and counter slang.
The part was there: the 244-case study
In parallel, we judged by hand 244 part-search cases across 5 auto parts stores (samples per store plus a full census in two of them, with the method audited). Two results that sting:
- The human counter found the correct part in 48% of the cases judged.
- In 3 out of 10 orders, NOBODY found the part: not the person, not the system. And when we checked against the catalog, in 69 of 70 cases reviewed the part WAS there. It is not missing inventory: the search fails (slang that does not connect, a cross-reference nobody runs, the rush).
That gap between "it is in the catalog" and "nobody finds it" is where the silent sale is lost: the customer never complains, they just buy somewhere else. It is the reason an AI quoting tool has to be evaluated against your real data and not against a demo: the difference is how many of THOSE it finds.
Traceability: nobody knows what happened to the quote
In the measured store's system almost none of the follow-up gets recorded: there is no way to tell whether the customer was quoted, bought, or went to the competition. With no traceability there is no follow-up, and with no follow-up whatever is left on read is invisible to the owner.
Citable figures from this study (CC BY 4.0)
- First price on WhatsApp: median 16 min (n=31,352 requests with a price); monthly median between 13.2 and 18.5 min across 20 months (n=31,352 requests with a price, Jan-2025 to Aug-2026).
- 1 in 5 requested parts go unquoted in the first 24 hours (20%, Jul-Aug 2026).
- Manual part search: 48% human hit rate; in 3 out of 10 orders nobody finds the part even though in 69 of 70 verified cases it IS in the catalog (n=244, 5 auto parts stores, audited).
- 29.6% of requests arrive with a photo and at least 95 out of every 100 arrive with no part number (n=57,681 requests, 20 months).
Time-to-price method: request = the customer first message after the previous loop closed; price = the counter first message containing a figure; capped at 24 hours. When citing: "Suplifai, study of auto parts quoting on WhatsApp, Mexico 2025-2026" with a link to this page.
What to do with these numbers (an honest checklist)
- Measure your time to PRICE, not your time to reply. The instant greeting hides the real delay. The question is: how long between the customer's message and the first price in the chat?
- Count your "left on read". How many of this week conversations ended with a customer message and no answer?
- Record the status of every quote (sent, won, lost). Without that, follow-up runs on memory.
- Test your process with photos and slang, because that is how the real customer asks: the one who quotes with a part number is the exception.
- If the bottleneck is finding the part, the fix is not answering the greeting faster: it is making the price arrive as fast as the reply does.
Slow to quote parts? The mechanic won't wait
The mechanic won't wait: they buy from the first store that gives them a price. In the operation we measured, the chat answered in 2 minutes and the price took 16 (medians over 20 months: 56,197 first replies for the 2 minutes and 31,352 requests that got a price for the 16). In those 14 minutes your people search and key in data instead of selling.
Suplifai reads the part number, checks your inventory and quotes on WhatsApp in seconds, 24/7, so your team spends its time closing sales instead of searching catalogs for prices. Message us and see a real quote live.
Also read: swamped with messages asking for auto parts and quote auto parts even when your store is closed.
Frequently asked questions
How long does an auto parts quote take on WhatsApp?
In the real operation we measured (20 months of WhatsApp): the chat replies in 2 minutes, but the first PRICE takes a median of 16 minutes (31,352 requests with a price), and 1 in 5 requested parts go unquoted within 24 hours (Jul-Aug 2026). The delay lives in finding the part, not in answering the customer.
Why do prices take so long if they reply so fast?
Because replying and quoting are two different jobs: the greeting and the 'let me check' go out instantly (often from canned replies), but identifying the part, cross-referencing the number, checking stock and putting the price together is still manual work squeezed in between customers.
How do auto parts customers ask for quotes today?
In plain language, in counter slang, and with PHOTOS: 1 in 3 requests arrives with a photo, and at least 95 out of every 100 arrive with no part number. A quoting process has to understand photos and slang, not just SKUs.
What is time-to-price?
The time between the customer's first message and the first price in the chat. It is the honest metric for a quoting operation: you can reply in 2 minutes and send the price hours later, and the customer buys where the price arrived, not where the greeting arrived.
How many sales do I lose by being slow to quote?
There is no money figure that fits every auto parts store, and we don't make one up. What we did measure: the price takes a median of 16 minutes to arrive, and 1 in 5 requested parts go unquoted within 24 hours (Jul-Aug 2026). Count how many of your own requests end up without a price and you'll have your own number.