Price tracking website shows you if a Takealot deal is actually good

[Jan]

This new website tells you if a Takealot deal is good

A new website makes it easy for Takealot shoppers to see whether they are scoring a good deal on a discounted product or whether the saving shown to them has been overstated.

Servaltracker.com records the daily historic listed and selling prices of popular products on South Africa's biggest ecommerce site and presents these in a simple graph.

 

[Other Pineapple Smurf]

Sadly this will be blocked by TAL.

 

[Gazg]

Interesting.......

 

[PhireSide]

huh, interesting.

I wonder whatever happened to PriceBoar, now that I think about it

 

[chrisc]

It does not matter if TAL like it or not. All you need is the URL

Generally speaking, TALs merchandise is at the cheap end of the market and much of the products are of poor quality

 

[PsyWulf]

It does not matter if TAL like it or not. All you need is the URL

Historically some sites have blocked price scraper/comparison sites when they get popular

 

[chrisc]

But how can they? TAL does not run the internet

 

[PsyWulf]

But how can they? TAL does not run the internet

From a technical standpoint?

Scraper uses bot/engine to scrape,this is hosted on IP 266.266.266.266
TA sees high volume of scrape from 266.266.266.266 and *blocks* it
Whelp sorry it broke

Hudson used the Python-based open-source task queue protocol Celery to create a simple “worker” that walks through paginated responses from Takealot’s API and scrape the pricing data.

 

[PhireSide]

But how can they? TAL does not run the internet

They probably implement CAPTCHA or IP blocking based upon the number of requests coming from a single IP address

 

[ToxicBunny]

From a technical standpoint?

Scraper uses bot/engine to scrape,this is hosted on IP 266.266.266.266
TA sees high volume of scrape from 266.266.266.266 and *blocks* it
Whelp sorry it broke

From that side... Easy enough, also easy to circumvent by just rotating the public ip of the worker...

Tal could also block it by implementing stuff that stops the worker from scraping the site, or that changes the URL randomly per product.

 

[PsyWulf]

From that side... Easy enough, also easy to circumvent by just rotating the public ip of the worker...

Tal could also block it by implementing stuff that stops the worker from scraping the site, or that changes the URL randomly per product.

Just 1 example for the exasperated chrisc suffice to say there is a litany of anti-scrape tech

 

[JustAsk]

...except that if Takealot, like many others, kept the price artificially high for 2/3 months before silly season, it will show you a fake bargain. Helping takealot with exactly what they wanted to achieve.

 

[Dark Agent]

Historically some sites have blocked price scraper/comparison sites when they get popular

We did something similar to isthereanydeal.com, we scrap of google caches, makro, games, takealot hi-fi, incrediable. The nice feature we had was machine learning, which group similar products together, like 9KG washing from Makro with a 9KG washing machine from game, nor different bar codes it will match base on feature sets. Worked great on 10000 products since API request was just daily. Get to 20000 and it will become malicious traffic.

 

[Peon]

Cease & Desist order coming in 3..2..1..

 

[Dark Agent]

...except that if Takealot, like many others, kept the price artificially high for 2/3 months before silly season, it will show you a fake bargain. Helping takealot with exactly what they wanted to achieve.

Tracking from one source is pointless, might only work for the daily deals.

 

[rvZA]

But how can they? TAL does not run the internet

A script is used to scrape daily data on Takealot. Easy to identify when this happens, easy to block IP Addresses and down goes the website data.

 

[konfab]

From a technical standpoint?

Scraper uses bot/engine to scrape,this is hosted on IP 266.266.266.266
TA sees high volume of scrape from 266.266.266.266 and *blocks* it
Whelp sorry it broke

Hence why all the successful price scraping apps use a browser extension to read page content. Only way takealot can stop that is to scrub their frontend of machine readable identifiable markers on products. But that isn't an easy task by any means.

If Takealot actually wanted the power move, it would be to allow scrapers, but identify them and record what they are monitoring.

Then they can use that to determine whether to run a sale on a particular product or not.

In fact, if they really wanted the power move it would be to do what Steam does and notify users when a product on a wishlist is on sale. That takes all the wind out of the sails for any sort of scraper because completely fills the need

That approach caters for both impulse buys and more cautious buys.

 

[Totempole]

This is brilliant. Been waiting for a service like this forever. Similar concept to steamdb for Steam games on sale.

 

[Fulcrum29]

 

[rvZA]

Pretty much this....