Reboot 2003

Chapter 28 028: DSP Platform
The so-called DSP platform is a "demand-side platform", which emerged in the 3G era when advertising targeting technology flourished, and represents the development trend of the advertising industry from "media-oriented" to "precision delivery".

Playing on the DSP platform in this era seems to be a bit ahead of its time.

However, this technology itself does not have a big technical threshold, and some small problems can also be solved by the traditional means of the advertising industry. Whether it can be popular and why no one has done it before probably starts with the pricing method of Internet advertising.

In 2003, Internet advertising had just emerged from the doubleclike era, and Google Ads had not yet emerged from its own private land. The advertising and marketing industry was basically dominated by media, so the pricing method was often dominated by media ideas.

For the media, the most important task is to maximize the total amount of traffic. As for how to sell the advertising space, they often don't care much.

Either it is time-based, that is, an advertising space is rented at a fixed price, and the media only needs to replace it according to the time, and just lie down and collect money.

Or use the number of impressions to calculate the price. An advertisement will be displayed to the user in units of thousands of times, and then the price will be calculated according to the final data.

However, these models are obviously unfair-

Not to mention the tricks and shady scenes that may exist, nor the series of disputes that may be caused by the opaque billing method.

More importantly, whether it is the display time or the number of impressions, there is a considerable distance from the needs of advertisers. They do not know how directly these data have a direct relationship with their final needs.

At this time, advertisers will need a demand-side platform.

The demand-side platform in the future will provide many pricing methods that are closer to the advertisers, and the extreme ones will even pay based on user behavior and final sales. These billing methods will allow traffic to be put into the market in a more clearly marked way, allowing media parties to pay more attention to advertising benefits.

It also promotes the progress of the industry through the trading method of bidding and buying, allowing advertisers to pay more for excellent advertising spaces. From the perspective of the game between the two sides, advertisers will be more confident and bold in paying, and traffic owners will find ways to make advertisements more attractive.

In this way, both parties can pay more and make the effect of advertising more impressive, which is the biggest "pain point" that the DSP platform can solve; and Chen Qin just wants to build a DSP platform in advance to get a share of the vast advertising market.

Of course, it is not realistic for him to build such a platform from scratch.

But even if you just improve the pricing method and build it into a platform, it will be enough to solve some problems——

Coincidentally, the market that Chen Qin was targeting before was the middle and low-end transactions in the advertising industry. Due to the obvious "long tail effect" in the field of online advertising, those small advertisements below 5000 yuan can actually provide very rich profits.

And these advertising spaces happen to be the hardest-hit areas of opaque pricing—not to mention the traffic owners, because they can’t find any customer resources at all, so they can only catch the advertisers who have finally come to them, and they fall into a vicious circle.

Advertisers have a little money or advertising space in their hands, but they can't spend it. They finally found some small advertising spaces in the corners, but in fact they have no effect at all.

If pricing can be done in a more effective way, then the advertising transaction will become more transparent, and both parties can get what they want. As for the pricing method that Chen Qin wants to choose, it is actually very simple, that is, pricing by click.

Googleads, the future industry leader, is priced by the number of clicks, that is, once a user clicks on an advertisement, it is billed once. This method is sandwiched between the needs of advertisers and traffic owners. It is a moderate enough option, and more importantly, it is simple enough——

There is no need for too complicated scripts, just add a layer of jump to the advertisement to achieve it.

On this basis, Chen Qin can build a platform, then build a statistical database, display the statistical data, and then slowly add bricks and tiles, and a DSP platform can almost be formed.

In recent days, Chen Qin has basically been doing this.

Of course, I am basically busy with this at night, writing logic maps in math books during class during the day.

The actual efficiency is not very high, but fortunately, he has the examples in his memory as a reference, and he has done both front-end and back-end, and then found some ready-made runtime libraries and frameworks to build them up. Today, his platform has almost a prototype——

This almost means that it can run, accounting for about 30% of the project.

And the next development can be iterated while debugging. In short, the station must be opened first.

To build a website requires a server, and most start-up websites can be opened with their own computers.But Chen Qin couldn't keep Xu Lingling's computer on all day, so he had to rent a server first...

Fortunately, renting the server itself is not very expensive. The dedicated host is at least a few hundred a month, but a virtual host is enough in the early stage, and it can be sold on godaddy.

In addition, starting a website also requires a domain name. Chen Qin doesn't have many other things, but there are really many domain names.

It's just that the English name of San Diego has been snatched away long ago, and it seems useless to spend money to buy such a domain name now.

Chen Qin searched in his collection before, those hot words are not easy to use here, in the ordinary Shuangpin that he registered later, he found kuaikan (quick look).com.

The analysis was successful tonight. Chen Qin went up to try it himself, entered the website address, and was able to enter the semi-finished site he made.

Chen Qin started messing around with debugging bugs. At this moment, someone on msn talked to him. It was Boss Liang from the private server.

"Expert, are you there?"

"I'm sorry, the business has been upgraded recently, and we won't accept advertising business for now."

There is no way, one person has limited energy, and if he wants to upgrade the industry, the original business can only be put on hold temporarily.

However, Boss Liang has always been a lively type. Seeing that Chen Qin replied manually, he chatted with him familiarly: "When will it be restored? I still think your advertisement is better!"

Chen Qin was curious: "Wouldn't it be good for Yewang?"

Because he can't get such a big share by himself, and he doesn't plan to rely on business to get the share, during this "retreat", these regular customers came to him, and Chen Qin basically sent them to the future competitor Haoye.com.

However, Boss Liang seems to have a lot of opinions on Haoye.com: "Go, I just think that Haoye.com is cheating people. My advertisement is completely useless, and it is not as useful as yours."

Chen Qin was curious: "Is Haoye.com not available for you?"

"Who knows, I just followed the model here and sent them the money to let them arrange it, but after half a month, it didn't work at all, and I thought it wasn't up yet." Boss Liang complained, "When I asked them, guess what. They said they would go up on the same day. I was fooled. This was a hammer. The 2 yuan advertisement came not as much as the 1 yuan advertisement. Thanks to me, I added a server in advance!"

(End of this chapter)