Facebook vs. F.B. Purity

ISO 9001:2015 clause 4.2—“Understanding the needs and expectations of interested parties” requires the organization to determine the “requirements of these interested parties that are relevant to the quality management system.” The recent two-day conflict that Facebook lost to F.B. (Fluff Busting) Purity and Adblock Plus illustrates the rare but important situation in which houses of quality contend much as the houses of Capulet and Montague feuded in Romeo and Juliet.

This latest conflict began when Facebook changed its computer code to make sponsored content indistinguishable from desirable content to get advertising past ad-blocking utilities. Facebook can make this work because its ads go through its own web domain, while third-party advertising domains can simply be blocked by a computer’s hosts file. In a Wall Stree Journal article, Andrew Bosworth, vice president of Facebook’s ads and business platform, proclaimed that, “Ads are a part of the Facebook experience; they’re not a tack on.” He doesn’t seem to understand that the customer rather than the supplier defines what constitutes a value-adding experience.

F.B. Purity then needed roughly 24 hours to announce, “F.B. Purity v. 15.1.0 has been released. It fixes the issues caused by Facebook’s recent code change that attempted to stop people blocking ads on the site, and filtering out the stuff they don’t want to see in the newsfeed.” The next day, Adblock Plus announced, “Facebook blockade is OVER! Facebook stopped letting you block ads two days ago. That was so two days ago. Read how to start blocking again here,” and directed the user to its website.

Facebook’s ability to compel its customers to view unwanted ads therefore lasted less than two days. Its executives are admittedly taking actions to identify risks and opportunities, even though their efforts to address them are merely aggravating the risks in question. In a USA Today article (“Facebook to block ad blockers on desktop”) Jessica Guynn writes:

“Facebook listed ad-blocking software as a risk in its most recent quarterly filing.

‘Revenue generated from the display of ads on personal computers has been impacted by these technologies from time to time,’ Facebook said in the filing. ‘As a result, these technologies have had an adverse effect on our financial results and, if such technologies continue to proliferate, in particular with respect to mobile platforms, our future financial results may be harmed.’ ”

This risk will only intensify as long as Facebook and other advertising-funded websites fail to reconcile, as well as identify, conflicting “needs and expectations of interested parties.” These websites can, on the other hand, implement advertising models that will ideally add value for users, or at least not alienate them enough to block the advertising or even leave the websites entirely.

The contending houses of quality

The traditional quality function deployment (QFD) model identifies and prioritizes the needs and expectations of the relevant interested parties, cross-references them with product or service features, and estimates the interaction between the customer requirements and the product features. The interaction matrix, or “roof” of the house of quality, identifies synergies (+) and antagonisms (–) between the product or service features.

Antagonisms are, however, tradeoffs between desirable product features such as low cost and high durability. They can’t adequately describe situations in which one interested party’s value is another’s muda, and we may actually have to break new ground with the introduction of a negative interrelationship rating as well as separate importance ratings for each interested party, as shown in figure 1.

Figure 1. QFD for conflicting needs and expectations

The first takeaway from this figure is that a separate importance rating is calculated for each stakeholder or interested party. This shows how QFD can be modified to address “needs and expectations of interested parties” when there are more than one. Even if needs and expectations don’t conflict as shown in figure 1, importance and/or interrelationship ratings may differ between interested parties.

Interrelationship ratings are normally on a 1 to 5 or 0 to 5 scale, and I’m using 0 for that between the advertising and the utility of the actual content. Advertising doesn’t affect the useful content itself, although it does affect its accessibility. The user’s primary concern is that the content be useful (5), and the secondary concern is that it be accessible; 4 is used in the example. The advertisers’ and Facebook’s primary concern is that the advertising be seen (5), and I’ve given the interrelationship between this requirement and the utility of the content a 4. Useful content brings more visitors to the website in question, so it’s important to both interested parties.

The fact that requirements come from two or more interested parties, however, results in a negative number for the interrelationship between ads and accessibility. This can range from 0 to negative 5, where 0 means the ads don’t affect accessibility, and –5 renders the product or service so undesirable that customers will stop using it unless they can block the ads. The fact that the sponsored content is placed directly in the middle of the user’s timeline, but doesn’t involve pop-ups or unauthorized video or audio, makes x roughly 3: large enough to annoy most users but not enough to drive them away from Facebook. I’m using 3 as an average because some people don’t care about the advertising at all, while others have stopped using Facebook because of it.

The importance rating for a product or service feature is the sum of the products of the customer priority ratings and the interrelationship weights. As an example, the user’s importance rating for content is 5 × 5 + 4 × 3 = 37 out of a possible 50, while that of the publisher and advertiser is 5 × 4 = 20 out of a possible 25. This shows that the actual content is important to all the interested parties. Noting, however, that 50 points are possible for the user, and only 25 for the publisher and advertiser, it makes sense to report percentages of the total possible points rather than total points, as shown in figure 2. This allows easy comparison of the relative importance of the product features by interested party. As an example, the importance of the content is 37/50 or 74 percent of the 50 possible points for the user, and 20/25 or 80 percent of the publisher’s possible 25 points.

Figure 2. House of quality with importance ratings as percent of possible total

The advertising is meanwhile of utmost importance (25 out of 25 or 100%) to the publisher and its advertisers, but because of the negative interrelationship with the accessibility—that is, the advertising interrupts the flow of the content—it is of negative utility to the user. This, in turn, creates a demand for another product feature that is emphatically not provided by Facebook, YouTube, or any other website that relies on advertising revenues. Figure 3 assumes that the interrelationship between the “content is accessible” requirement and the ad blocker(s) has the same absolute value as that between “content is accessible” and the advertising. The interrelationship between “advertising is seen” and the ad blocker is, of course, –5.

Figure 3. House of quality with third-party product feature

The negative interrelationship between “content is accessible” and “advertising” creates the demand for ad blockers, and the only way to remove this demand is to reduce this negative interrelationship to zero, or at worst, –1. Adblock Plus reports that 25 percent of users will block all advertising regardless of its placement or presentation, which suggests that zero is not attainable. According to Adblock Plus, a –1 (i.e., 75% of users will not block the advertising) is achievable by noninvasive placement of the ads. If Facebook were to place ads in a noninvasive manner, this would, as much as possible, reconcile the conflict between the interested parties’ houses of quality.

It’s difficult to think of situations other than advertising with conflicting houses of quality. As an example, movie DVDs with unskippable ads are likely to be given poor quality ratings, and rightly so. “Disk does not respond to fast forward or skip commands” is valid cause for a 1 out of 5 rating at Amazon and elsewhere, regardless of the DVD’s actual content; the purchaser is under no obligation to excuse poor quality regardless of the reason. The same goes for the Flash player’s failure to respond to a fast-forward command; “controls don't work” also is cause for a bad performance review regardless of the reason.

Another example of conflicting houses of quality can, however, be presented for variety. Government regulatory agencies such as the Environmental Protection Agency (EPA) set standards for vehicle exhaust emissions that affect public health. This makes the public as well as the EPA a relevant interested party. Many vehicle purchasers, on the other hand, assign a negative importance rating to emission controls that reduce the vehicle’s performance. This may have prompted Volkswagen’s misguided efforts to defeat emission tests with software that reduced the car’s performance only when it was under test (Hotten, Russell. "Volkswagen: The scandal explained." BBC News, December 10 2015.)

Abraham Lincoln pointed out long ago that a house divided against itself cannot stand, and this certainly includes a house of quality in which one interested party’s critical need is another’s defect. Pull advertising or customer-driven advertising may address this issue to generate a win-win outcome.

Pull or customer-driven advertising

Definitions of pull advertising vary, but we’ll define it similarly to pull production systems such as kanban and drum-buffer-rope. Pull production systems produce nothing for which there is no need, and pull advertising does not advertise anything for which there is no demand. Pull advertising performs a value-adding service by helping the user find what she needs. I found F.B. Purity, for example, through a Google search on “block Facebook ads” when invasive ads showed up in my browser.

F.B. Purity’s pull advertising model—people come looking for it and similar products with no advertising expenditures whatsoever—is enormously successful. Suppose, for example, that only 10 percent of Facebook’s desktop computer users consider the sponsored content invasive and, of those, only 10 percent make the recommended payment to F.B. Purity. One percent of hundreds of millions of people still adds up to millions of paying customers. The takeaway is that people will generally look for solutions to their problems rather than wait for solutions to come to them.

This underscores the difference between push advertisements that get into customer’s faces with little if any idea as to whether the customer wants the item in question, and a situation in which a customer comes looking for a product or a solution to a problem. I don’t know what fractions of Google Shopping, Amazon, and eBay searches result in sales, but they must far exceed, probably by orders of magnitude, the fractions of ad impressions that result in sales. Consider the following approaches to advertising as shown in figure 4. A sale results when our product or service matches the customer’s needs, and the chance of a sale is the ratio of the area covered by our product features to the area over which we spread our advertising.

Figure 4. Generalized, targeted, and pull advertising

• Generalized ads are published with little effort to identify the customer’s needs. The ad is thrown out with the hope that somebody will buy, much as aircraft dropped thousands of bombs during the WW II with the hope that at least a few might land on something important like a weapon factory or oil refinery. This approach is also similar to “spray and pray,” a bankrupt gun-fighting tactic in which one fires as much ammunition as possible in the enemy’s general direction to increase what reliability engineers would call the enemy’s hazard rate from infinitesimal to at least negligible.

• Targeted ads are directed at consumers through tracking cookies on the internet, and by likely viewers of television shows. Kitchen products might, for example, be advertised on cooking shows. Internet advertisers, however, face the issue of privacy controls such as “self-destructing cookies” because many users don’t want unknown entities to be able to track their browsing habits. However, the latter add-on can safe-list or whitelist cookies from websites such as asq.org, qualitydigest.com, aiag.org, and so on.

• The customer-initiated search looks for products or services similar to ours, which not only makes a match far more likely than for any targeted ad, but also involves somebody who is at least thinking of buying. This is obviously the best situation.

Note also that, in the cases of generalized and targeted advertising, the advertiser must pay to cover (carpet-bomb or spray-and-pray in the case of the generalized ad) the area in question while, in the case of pull or search-driven ads, potential customers target themselves on what we are trying to sell.

Facebook does have at least a rudimentary pull advertising capability through its search feature. As an example, a search on “ISO 9001:2015” delivers, on the initial page of the search results, contact information for at least two firms that sell consulting and training services relevant to ISO 9001. Searches on “metrology” and “calibration” are equally instructive. The second search generates, again on the first page of the search results, two calibration laboratories with ISO 17025 accreditation. A search on “ASQ certification” leads to a company that offers training not only for ASQ’s certification exams but also for PMI’s Project Management Professional and APICS’s Certified Supply Chain Professional. Facebook searches can also include geographic considerations: a search on “home,” “inspection,” and “Philadelphia” identifies home inspectors in the Philadelphia area. This is also true, of course, on any Google search.

However, as far as I know, Facebook doesn’t charge these businesses anything for clicks on their pages, although it does charge them to boost their posts (i.e., advertise them to other users). Amazon and eBay, on the other hand, earn money when people find and purchase products. This suggests that Facebook may actually be overlooking a pull-driven advertising and sales model, and vendors might easily be willing to pay commissions to Facebook on a per-sale basis.

About The Author