13 January 2010

Targeting a Conglomerate Modeled AQ

How can we more effectively target AQ through the use of a conglomerate model?

As we’ve seen over the course of eight years, AQ gets an “A” for resiliency. Combating the organization has proven far more difficult than probably originally anticipated after 9/11. Continued attempts to lump all subordinate groups into one homogenous entity are counter-productive to long-term efforts; it makes US and/or UN efforts and operations in one Theater appear futile because the overall organization still appears strong to the global audience. In essence, successful efforts in one Theater are immediately overshadowed in the eyes of the global audience by an attack executed by another group in a different Theater or region. “Sure they picked up a bunch of terrorists in Iraq, but the Afghan Taliban did this, and HSM did that.” In this final section I will offer lethal and non-lethal concepts for targeting a conglomerate-modeled AQ that, I believe, will benefit our global endeavors.


Using my slightly altered definition for conglomerate, where I replaced “unrelated industries” with “regions”, a conglomerate model allows for a better conceptual strategic targeting framework. In a franchise model, victory will only come through the defeat of the corporate headquarters. Individual franchisees will board the doors if their higher HQ goes under; you won’t see an independent Jimmy John’s if corporate HQ fails. In a conglomerate model, the higher headquarters can be targeted through its smaller companies.

Most importantly, a conglomerate model better incorporates national and regional dynamics in play. These dynamics help tailor national (or international) responses to combating these subordinate terror groups. US support to both Ethiopia and the Somali TFG, and direct support to the Yemeni government, are great examples of national and regional dynamics affecting targeting. Though Ethiopia failed in their AMISOM mission, their role in East Africa cannot be ignored. In Yemen, the West has chosen to provide support primarily to the Yemeni government, and not outwardly support Saudi incursions south of the border.
Depending on the scope, a conglomerate model suggests it is more important to target intermediaries and subordinate terror group leadership. AQ Central Leadership becomes less important in this model. I am not advocating that we stop our search for those at the pinnacle of AQ, instead I advocate placing personalities like Hakimullah Mehsud, Sheikh Ahmed Abdi Godane and Abu Saleh al Somali (EKIA 8DEC08) on an equal tier. Personalities like Abu Saleh al Somali are especially important; they not only serve in a “senior conglomerate” staff role, but also serve as a critical node between HSM and AQ Central Leadership. These critical nodes are just as important in a conglomerate model as they are in a flatter network model, and are essential to successful targeting.

A conglomerate model will help intelligence analysts identify terror groups attempting to align under AQCL earlier. Multiple groups have stepped up attacks outside their borders. AQAP was nearly successful in their attempted airline attack; AQIM continues to target and kidnap Western tourists across the Sahel; HSM conducts attacks in bordering nations Ethiopia and Kenya, and is also widely believed as ultimately responsible for the recent attack on Kurt Westergaard. Hamas was snubbed because they were not “continental” enough in their terror undertakings. Another group in another region will again attempt to align by taking their fight regional or more. This “escalation” could potentially serve as an early indicator (or key milestone for fledgling salafi-jihadist groups) for the intelligence community to provide additional analytical effort.

A conglomerate model (especially vs. a franchise model) also suggests there is no cookie-cutter solution to targeting and combating “subordinate” terror groups and allows a framework for more innovative targeting methods or techniques.


Marc Lynch’s 11 JAN article on Foreign Policy covers the over-arching theme of my non-lethal targeting recommendations. The term “GWOT” is not only incorrect, it is counter-productive. I believe much of these generalizations draws from a conceptual (mis)understanding of AQ as a franchise-style organization. A conglomerate model allows Western and Host Nation governments to more effectively prosecute a longer-term information operations campaign. Principally, a conglomerate model reduces our threshold for success, much in the same way AQ elements have done for quite a while now. Success in one region against one conglomerate subordinate becomes just that, success. By removing the global context, we remove the argument of AQ as an unstoppable movement. This provides us with a great opportunity to marginalize AQCL. The emphasis is placed on subordinate groups with their own problems, issues, complaints, etc. These central problems become the focal point.

By de-emphasizing AQ as any kind of united or homogenous entity, instead marketing them as disparate regional or national insurgencies and/or terror groups, we set the conditions for fissures to develop between AQCL and subordinate groups. The 30 DEC Khost attack is a prime example of a fissure that we could have exploited. Three separate entities later claimed responsibility for the attack: the Afghan Taliban, AQ and TTP (Pakistani Taliban). After the apparent foolishness of AT and AQ’s false claims, a conglomerate model would have highlighted the potential fissure and allowed us to attack AQ and subordinate groups with tailored messages. In terms of AQCL we could have highlighted the unwillingness of subordinate groups to cooperate with AQ’s attempted media plan after the strike. By continuing along this line, we would hope to create an over-reaction on the part of AQCL by attempting to rein subordinate groups back in. Toward subordinate groups we could have highlighted that AQCL wants to steal the "glory" that should be bestowed upon the small group. Taking it one step further, we could then have pushed for recognition that AQCL is not actually in charge of each group’s destiny, they drive their own train and fight their own fight. The non-lethal targeting opportunities after the IO fumbling were tremendous; sadly, from my vantage point we missed a great opportunity.

The hope obviously in all of this is that these fissures would grow and eventually cause AQCL’s own subordinate groups to marginalize AQCL and thus weaken the accompanying brand/ideology.


  1. Josh,

    Your analysis is thorough, and I agree with your expansive use of a business model. I read it as a metaphor, a pane to view AQ.

    The main question I continue to ask myself is: "What does AQ want to be?" They do have a centralized framework, but I think you are right to say that they operate differently in conglomerates. I think the hierarchical structure remains, though, in funding and approvals/blessings.

    What I am really asking is: "Does AQ want to grow, and what does that line of growth look like?" Do they operate like a conglomerate now but plan to advance to a franchise later? Is franchise higher on the chain than conglomerate, or should we accept that they are happy with where they are at?

    This is why I think Dr. Ronfeldt's commentary is precisely important, as well as the theories that have been visited before. It calls us to question again the tribal definition, because I think the tribe paradigm limits the importance of structured hierarchy and/or their willingness for strategic growth.

    In my 8 January post, I stated;
    "[AQ] innovation may be occurring at the organizational level, but we are focusing on the entity level only; i.e. growth of Al Shabaab. Groups may be beginning to talk intra-organizationally. This would cause a systematic shift in ideology, tactics, strategy, and operations. The shift need not be small but only effective; it need not be complex, but only coordinated."

    The term "intra-organization" communication highlights AQ as a network, that which Pat has added to the argument. AQ affiliates could be operating as a conglomerate as you say - applying their strengths - within the larger network. In a network, the main importance is that particular actions of one must benefit the universal connection of all. This would further support a franchise element or the development thereof.

    Just some thoughts I wanted to pass on to discuss the way AQ may be acting intelligently. Basic science tells us that every living cell naturally yearns to grow, and together we can analyze how and why AQ would want to do so to be effective according to their "mission objectives" and "operating procedures."

  2. I see some value in evaluating AQ relationships with other organizations, in the same way that we analyze various types of conglomerate strategies. For example, there is a good body of research that suggests conglomerates that diversify into related industries have greater success than those that diversify into unrelated industries. Some analysis gets into more nuanced categories than just related or unrelated. Examining why certain conglomeration strategies create less value may help to shed some light on potential fissures if AQ exhibits behavior consistent with a conglomerate.

    However, I question your second-to-last paragraph. The targeting recommendation that you outline and attribute to a conglomerate analysis happens to be almost exactly how we targeted fissures between member organizations within the "Islamic State of Iraq" in 2007. We pursued that targeting even though we were working on a franchise model (I don't know if the franchise model was uniformly accepted, but most S-2 personnel that I knew embraced it). I'm not so sure of your apparent assertion of bad targeting for want of this model.

    I would also question the applicable scope. You are applying this model at the transnational level and at the level of country, province, and city. It seems a stretch to think that it holds true for all of them. Suppose that AQ is the conglomerate, AQI is a subsidiary, and there are "lower tiers" in Ninewa and "tiers" that are lower still in Mosul. What is the significance of viewing AQ as a conglomerate when conducting operations against AQI operatives in Mosul? At this level, the brand image (I'm not arguing franchise here) seems more relevant. I specifically remember reading the complaints of Mosul residents on internet bullet boards who were angry at local gangs whose members pointed guns at people and demanded money, claiming, "I'm waging jihad, give me your money." The residents griped that, "these men are not jihadis. They are just criminals claiming to be jihadis." AQI's brand image suffered from that because any real operative needed to constantly defend his organization against the frauds telling lies about jihad to justify their thievery.

    You may also want to consider other concepts, such as joint ventures. When I think about the relationship between the Afghan Taliban, AQ, and TTP in your second-to-last paragraph, I don't think conglomerate or franchise. I think about the relationship that Dunkin Donuts has with various grocery stores, banks, and gas stations; or the relationship between Target and Starbucks.

    I wrote something similar in your original thread on this topic, but I'll try to reiterate more clearly here: consider also that conglomerate and franchise are not mutually exclusive. I think a good argument can be made that AQ, on the transnational level, is a conglomerate. But its Iraq subsidiary, AQI, was a franchise. It's subsidiaries in Yemen, Somalia, or elsewhere might be better described as greenfield ventures or acquisitions (both consistent with the conglomerate model), but those business units, within their "market," might pursue a franchise operation or something else.

    Lastly, in a previous post you typed, "In a network, you see a 'flattened' method for decision making, communication and execution. Persons inside a network can come and go as they please, and can connect freely with anyone else inside that network."

    I think that goes too far. Perhaps in a "completely flat" network, that would hold true. But in an organization that is simply "flatter" (but not completely so), there is not really a come and go as you please freedom. Members of the network may travel to other parts of the network, but not until told to do so (good example would be al-Badri moving from Samarra to Baghdad in 2005 - he neither invited himself, nor got an invitation from Baghdad; he was directed there by the higher tier).

  3. Schmedlap,

    I wont address your entire response since it is not my post, but I did want to say a couple of things. I think analysis on the difference between the tactical, operational, and strategic levels of AQ with respect to Josh's Conglomerate model gives all of us more to think about.

    We may need to view AQ in Tiers that lend to conglomerate/franchise/network model. I am subscriber to the conglomerate model at the strategic level. I think it will help us develop a "High Payoff Target List" that targets the fissures that Josh mentions. I also agree with you that on the lowest tactical level a different network model could facilitate a better understanding of AQ and help us foresee changes in the group. You mentioned how al-Badri was directed to change AO's proving a flat network model ineffective. I would classify that as an operational level decision. On the tactical level, military emir and below, the flatter network model does fit in my experience. I witnessed undirected shifts of groups based on changes in the tactical situation. All and all, I think we need continue our analysis of this topic.

  4. "You mentioned how al-Badri was directed to change AO's proving a flat network model ineffective. I would classify that as an operational level decision. On the tactical level, military emir and below, the flatter network model does fit in my experience."

    I wasn't trying to prove a flat network model ineffective. I was just taking issue with the narrow point that members of a network can come and go as they please and freely connect with others in the network. My impression may be wrong, but it seemed as though he was defining a network as always perfectly flat (I may be misinterpreting). My assertion was simply that AQI was "flatter" but not completely so - members didn't just travel to new areas and join new networks on their own accord. It had characteristics of a network and of a hierarchical model - kind of like a lot of modern businesses. On the local level, perhaps a flat network holds true, and on the provincial level or higher a less flat "network of networks" model. I'll stop now - I'm way out of my area of expertise.

  5. Schmedlap,
    You make some viable points, as does JD. The original intent of this series was two-fold: the first was to dispel the notion held by many that AQ subordinates are franchises...at the highest level. I still feel this is incorrect and creates incorrect beliefs and manifestations by suggesting "cookie-cutter" solutions. The franchise model is being used globally now. People understand business modeling, it is easier to digest. My original post discussed AQCL and subordinate/affiliated/whatever you want to call them organizations like AQI, AQAP, HSM, etc. Certainly each organization has unique constructs. HSM has far more political/government type of guys in their organization than AQI or AQAP. You see where I'm going with this. What likely makes them all similar is their relationship with AQCL. This offers itself more to conglomerate than franchise at that level.

    The second point was to generate thought and discussion, for readers and myself.

    Much of my recent thought about networks has been shaped by the book "The Starfish and the Spider" by Ori Brafman and Rod Beckstrom. That original comment I made is a huge summary of the concepts they present in the book. Not sure if you've read it, but it's a worthwhile read. I will offer that not all networks are flat, but I think a pure network is. Also, we did see AQI members at the lowest levels abandon one cell leader for another; this is largely what you were getting at in your last comment I believe.