Nerdery and the Trouble With Seventy Websites

Seventy websites are an impressive possession until somebody has to update them. Then they become seventy opportunities for a perfectly ordinary Tuesday to go wrong. Purina had accumulated a global collection of brand sites built on different technologies. The brands had things to say. The machinery was making them wait.

Nerdery, the Minnesota digital consultancy, helped move those sites onto a common publishing platform. The interesting part is the apparent modesty of the idea. A company hired software specialists, and one of the useful things those specialists did was reduce the number of separate technical problems the company had to solve.

01 / The seventy-site problem

In Nerdery’s account, Purina’s unmaintained legacy solutions slowed time to market and made scaling difficult. The replacement used Drupal, a content management system, with Acquia Cloud Site Factory. Nerdery reports migrating 70 sites in five months and reducing security scans by 95%. Those are project-specific figures, rather than a promise for the next customer.

Publishing could become a marketing task instead of a recurring request for development support. Behind that change sat an organizational decision: a product owner and centralized content production teams could coordinate priorities, compliance and accessibility. The platform gave that arrangement somewhere to live.

A technical essay about the work describes a “contrib-first” approach: assess Drupal’s community-contributed components before commissioning custom development. That is an agreeable philosophy for a company paid to write software. Some of the best engineering judgment consists of knowing which code somebody else has already written well enough.

For a reader responsible for a cluttered technology estate, the lesson is specific. Look for work repeated across brands or departments. Decide who owns the shared rules. Reserve custom effort for requirements that actually deserve it. A common platform cannot settle a dispute over who gets to publish what. It can make an agreed answer cheaper to repeat.

02 / A consultancy with pocket-change origins

The name suggests a club that might argue about keyboards over lunch. The origin was a business gamble. Mike Schmidt, Mike Derheim and Luke Bucklin worked together at Minnesota Internet before starting Sierra Bravo in 2003. Twin Cities Business reported that they pooled $153 in pocket change. The business adopted the Nerdery name in 2011.

The same reporting described an unusual distribution of status: employees carried the title co-president. Autonomy was part of the founders’ account of how the place worked. Bucklin died in a plane crash in 2010; the company continued under his colleagues. Schmidt, a co-founder and former CIO, became CEO in 2020.

There is a business argument beneath the eccentric title. A consultancy sells the judgment of people who often understand a technical problem better than the person supervising the contract. Give those specialists room to speak, and the client has a better chance of hearing an inconvenient fact early. The title alone proves nothing; the permission behind it matters.

03 / Buying a team, buying a result

Today Nerdery’s menu covers custom applications, user research and design, data engineering and analytics, AI development, and cloud modernization. A customer might need help moving an application, connecting data sources, or making a digital service easier to use. These jobs often overlap. A handsome interface can still be held hostage by a database.

Its commercial model offers two straightforward arrangements. Under “with you,” Nerdery adds expertise to the client’s team. Under “for you,” it takes on fuller responsibility for strategy, delivery and optimization. Onshore and nearshore staffing give it different ways to assemble the work. Fees are scoped through a price consultation; the choice starts with the job and the responsibility the buyer wants to hand over.

That puts Nerdery in the market for digital consulting and custom software delivery. The real alternatives include hiring internally, adapting existing software, or engaging another consultancy. A buyer should compare the people assigned, relevant project experience, delivery responsibilities and ownership of the finished work. A service catalogue can make several firms look identical. The proposed working relationship usually cannot.

“These Are Never Final”A heading in Nerdery’s public Objective-C coding standards

The old public coding guide supplies a pleasingly small clue about its craft: standards are meant to evolve and be challenged. That is a useful attitude to bring to enterprise software, where a rule that once prevented mistakes can eventually become the mistake.

04 / The dealer belongs in the checkout

Consider Polaris. Expanding online sales can unsettle a manufacturer’s dealers, who may see the new shopping channel as a competitor. Polaris wanted a better fulfillment experience and stronger dealer relationships. Nerdery worked on ship-to-store capability that made the dealer network part of the online proposition.

The work examined the experience of customers, warehouse workers, call-center staff and dealership teams. It included a roadmap, an interface that could be piloted while the site stayed live, and clearer customer communications. Nerdery reports a launch in under five months and delivery times reduced by more than half.

The revealing choice was whose problem counted. Optimizing only the shopper’s screen would have missed the network behind it. The copyable practice is to map the people required to fulfill the promise before drawing the promise on a page. It applies just as well to a hospital appointment or an industrial spare part.

05 / A few hours, a great many cards

At Real Sportscards, the unit of improvement was wonderfully tangible: preparation for livestream card breaks. The retailer had tested its business on a third-party platform and wanted an independent one. Nerdery says it conceived, designed, prototyped and launched the Uncommon App in six weeks.

The published case estimates 10-20% cost efficiencies, reports preparation falling to 30 minutes, and describes four daily labor hours becoming revenue-generating livestream time. Estimated savings should stay estimated. Even so, the mechanism is easy to understand: hours spent arranging the event could become hours spent hosting it.

Google Cloud’s Next conference presents a different scale of coordination. Nerdery has worked on its web and mobile experience since 2017, including the switch to virtual in 2020 and subsequent hybrid experiences. Booking sessions, navigating speakers and delivering streams are ordinary features until thousands of people need them together. Google is both a customer and a technology partner.

06 / The recommendation needs a reason

A separate, anonymized pet-food case shows Nerdery’s AI approach. The recommendation system already considered information about an animal; its explanation was too generic. Nerdery combined proprietary expert logic with Claude 3 Sonnet to produce more personal explanations, working with expert and legal teams on prompts.

The company reports a production-ready proof of concept in three months and a 17% rise in purchase intent during the first two months after launch, measured through more “Buy Now” clicks. That measures an action preceding a purchase. It should not be silently promoted into a 17% revenue gain.

The broader lesson is an editorial inference: an explanation layer is valuable when the underlying recommendation deserves confidence. Better prose cannot repair bad expert logic. A team copying the idea needs review, a defined use case and a measure that distinguishes curiosity from completed transactions.

07 / Where the next contract might lead

Nerdery’s newer offerings include Civic Analytics AI, a Google Cloud-based platform bringing agency data into a hub-and-spoke architecture. In April 2026, it announced a Google Cloud Partner of the Year award in the education category. The recognized work included a Gemini-powered Virtual Teacher’s Assistant developed with Google’s Rapid Innovation Team, designed to ground conversational help in course content.

The portfolio has widened, but its most instructive examples remain concrete. A publisher stops maintaining separate foundations. A dealer becomes part of fulfillment. A card retailer gets its afternoon back. Each begins with a business activity that software is making unnecessarily difficult.

For prospective customers, that is a useful starting brief: name the activity, count the delay, identify the people who must agree, and decide what success would look like. Bring that to a consultancy and the conversation has somewhere to go. Bring only a desire to be more digital, and even seventy websites might not be enough.