All of us Google-search users know that it is free for us and profitable from advertising. Most of us don't know that the advertising is handled by a subsidiary, AdWords. Google-AdWords has now teamed up with Salesforce.com to follow up on ad responses by keeping track of sales leads. Much of the communication involved uses Google mail and can use the Google online word processor and spreadsheet.
The Economist describes the relationship thusly link here: "Like all AdWords customers, they can then choose keywords ("car repairs", say) and bid to have small text links displayed next to the results of any web search for that term. They pay only when users click on the advertisement and are taken to the advertiser's website. At that point Salesforce's service kicks in [if they are among its subscribers], collecting information about the user which then pops up on the Salesforce page of the advertiser's sales team, allowing them to follow up and sell something."
Speculation arises that success will be followed by a buyout. Another natural monopoly challenges Microsoft's.
This Salesforce thing is creeping me out. "Allowing them to follow up and sell something" sounds a lot more active than the user follows an Adwords link and sees the static Web pages of the advertiser, where they may (or may not) put something in a virtual shopping cart and purchase it at their leisure.
It sounds more like the user follows an Adwords link and then has to abandon yet another e-mail address due to spam, or worse starts getting circulars from the advertiser in their snail mail, which they can't dodge nearly so easily as moving house costs hundreds of dollars versus nearly nil to change e-mail address.
Pushy salesmen actually banging on the door would be even worse of course...
This has significant privacy implications, as well as the obvious spam implications of any kind of Internet-click-initiated "sales force follows up actively trying to sell something". A click on a text ad (or any other online ad) should not in and of itself be taken as soliciting further, more active advertising efforts such as e-mail spam, nor as authorizing the disclosure of personal information, nor as establishing any kind of business relationship. (I wouldn't even want an outright sale to be taken as soliciting active advertising efforts. Don't call me, I'll call you if and when I want another product of yours.)
One privacy issue raised extends to safety. The obvious bogeyman scenario is that a pedophile sets up a Web site and subscribes to this Salesforce service, then takes out Adwords on popular toy brands and the like. Kids or parents that surf and click the links unwittingly pop up a detailed biography (whatever data Salesforce gathers) on the pedophile's computer screen. If the parents, or the kid with parental guidance, previously bought something elsewhere and their shipping address was entered into the forms, and "elsewhere" also subscribed to the Salesforce service, the pedophile knows where he or she lives.
Oops.
Previously the creep would have to set up a legit-looking online toystore that strung people along and looked professional enough long enough for kids or parents to get to the shipping-address part of the checkout process to get that. Now they can let some genuine e-toystore do all that; all they need to put up themselves is a landing page mocked up to look like a domain-not-found error or similar, plus Adwords, plus the Salesforce service.
I suggest a trick that won't work with email if you don't own your own domain: if you ever supply a street address on a web form, and you have a house rather than an apartment, put a bogus "Apt. 101" or similar in the address, different for each place you give your address. The deliveries (when they don't just take your money and quietly disappear off the 'net instead, that is) should still arrive, and the shipping labels should reveal which site shipped it by which apartment number is on the label. Moreover, unsolicited bulk snail-mail with an apartment number is a dead giveaway as to which site started spamming you or sold your mailing address. Unfortunately, a pedophile showing up isn't so likely to reveal which apartment number they think your kids live in...and this also won't work if the next shipping address variation keeps being propagated as an "update" to all the e-vendors through Salesforce.
It might be time to start investing in a P.O. box ... or maybe to stop doing any e-commerce until all the privacy and fraud problems are worked out. I'm still waiting for them to develop a "smartcard" solution to replace bank and credit cards that embodies a private key from a public key cryptosystem key-pair and uses it in a challenge-response authentication scheme, such that the private key (needed to authorize each individual charge) is never exposed to any third party... (The card would be revocable if stolen, of course; the reader would be a little $20 gadget with a card slot at one end and a USB cable coming out the other, usable with any computer, your own or a Point of Sale device in a store or wherever. Better yet, the "card" would be more like a pendrive, and have a jack on it to accept a standardized cable. It would have a small LCD screen and OK and Cancel buttons, and maybe other functions such as to set/change and enter a PIN, which would slow down its use by anyone who stole it long enough for revocation by the bank after you reported it stolen to take effect. You'd connect it by a double-ended USB cable to your PC at home for ecommerce, or at the (bricks-and-mortar) store plug it into the end of a cable marked "attach iWallet here" or suchlike at the checkout. The device displays the amount of the transaction and a minus sign if it's going to deplete your funds. You maybe enter a PIN, then OK, or hit Cancel if you don't like the amount. Each individual charge requires the bank to receive an electronic transaction authorization consisting of the amount, the customer account number, the recipient (or donor if receiving a deposit), and a digitally signed hash of above using your device's embedded private key. Obviously the device sends to the attached cable this authorization, with the amount it displayed in the window the amount in the authorization, and the amount being one of the things hashed with the private key. The bank decodes with the public key and verifies the amount that was hashed matches the amount in plain. This makes it nearly mathematically impossible for you to be charged any amount other than what you saw in (your own!) device's window and clicked OK to. It also would include (and hash) a time stamp or transaction GUID or both, making it impossible for the store or whoever to record the transaction authorizing packet and then resend multiple copies to double-dip, or use it to recurrently charge you without you having to authorize separately each recurrent charge. No more sneaky "evergreen" or uncancelable AOL subscriptions...only the bank can now sneakily ding you on the quiet. OK, so it's not QUITE perfect. But it would take a couple of kids in a garage workshop and $10,000 to engineer the first working system, if that, and the "iWallet" or whatever it gets called would cost all of $20 to manufacture, and be much more universally usable and secure than a real wallet costing $40. It would double as a proof of identity -- exposed only if you accept a transaction, including a $0.00 dummy transaction used purely for ID. If the bank kept some info like date of birth, as it likely already does, this gives you an age verification thing. With a little more services provided via the bank, which also becomes a more general identity-and-authentication-provider service, it can double as driver's license or other ID as needed. Including photo ID, if those using the system can pull up a thumbnail on their laptop after you OK something with your iWallet. All of this for cheap R&D and very cheap hardware and potentially-free software. Why has it not already been developed? Given the rampant privacy, fraud, and other problems occurring online and sometimes off, it's rather strange. Or is there a conspiracy by those who LIKE that we have to use insecure methods of authentication? Under this scheme, banks could even provide trusted pseudonymous identities, or even temporary ones, with a fixed amount of funds. If they know the same person has a driver's license or is over 18 or whatever those authorizations can be included, but it can't function as photo ID obviously. The bank would still be able to out your pseudonym, under subpoena or whatever, but it would deter casual privacy problems by letting you use funds without then being hounded by a salesforce, or a pedophile, or whatever.
Of course, what stops vendors that get your shipping address from spreading the info around? Perhaps an obfuscated shipping address solution also needs to exist -- an auto-forwarding P.O. box. A cheap service gives you a P.O. box address, but just forwards everything received to an address given by the box owner on setting up the account; no physical box is actually needed, aside from a back-room holding area of some kind. The sender doesn't know your real address; the service provider won't reveal it to third parties except under court order. Using it for all online transactions would greatly reduce your risk of ending up stalked, and the service provider would let you change the P.O. box address cheaply or for free to dodge any spam you began to receive, with the old address no longer being valid after a month, say. (Parcels would still be delivered for the old address for maybe a whole year, but not plain letters.)
