Skip to content
NerdVision

Creds

Hey there logo hunter, you made it!

The logos tell you where our team has been. The work tells you how we think. Explore the decisions, measurement and results behind our client programs, then meet the brands our senior team has worked with throughout their careers.

Our team's career experience

Enterprise brands. Senior experience.

Capital Group logo
01

WHAT WE DID FOR CAPITAL GROUP/AMERICAN FUNDS

Strategy, planning and execution for paid search.

ProShares logo
02

WHAT WE DID FOR PROSHARES

Strategy, planning and execution for paid search.

IBM logo
03

WHAT WE DID FOR IBM

Strategy and planning for global buying including launching Watson.

Microsoft logo
04

WHAT WE DID FOR MICROSOFT

Database management and distribution, beta testing and database queries.

Kayak logo
05

WHAT WE DID FOR KAYAK

Programmatic strategy and execution in US, CA, EU and APAC.

Amazon logo
06

WHAT WE DID FOR AMAZON

AI Max and Demand Gen campaigns for Amazon Prime.

Mastercard logo
07

WHAT WE DID FOR MASTERCARD

Operations, technical set-up and back-end performance analysis.

Squarespace logo
08

WHAT WE DID FOR SQUARESPACE

Programmatic lead generation strategy and execution of SuperBowl campaigns.

The RealReal logo
09

WHAT WE DID FOR THE REALREAL

Social Amplification of Seasonal Sales

Booking logo
10

WHAT WE DID FOR BOOKING

Social Amplification for Travel Offers.

Panasonic logo
11

WHAT WE DID FOR PANASONIC

Brand and extensive video execution for B2B portfolio.

Case studies

01 / Capital Group

Promoting the Vodcast on YouTube

The challenge

Capital Group's Vodcast lived as a series, not as a single video. There was a playlist for the series and individual videos for individual episodes. The media job was to distribute both levels at once: keep the series discoverable, push specific episodes when they launched, and keep every route measurable enough to separate one audience from another. Over time the job had sharpened to a single headline KPI: growing the channel's subscriber base, priced per line, per week.

What we did

We built the program in layers. Discovery campaigns pointed to the series playlist on YouTube, with separate campaign rows for different audience routes and a unique tracking ID on each one. Episode campaigns then supported individual releases. The trafficking files for Episodes 2 through 6 show episode specific YouTube destinations, audience splits that included first party, retargeting, combined first party and retargeting, and YouTube retargeting, and a unique nv_yt tracking ID on every route. A separate YouTube Audio line ran alongside the video work, and Shorts joined the mix in August.

When the reporting shifted from views to subscriptions, the management changed with it. Our weekly pacing notes tracked subscriptions and cost per subscriber separately for Series, Episodes, Discovery, Shorts, Audio and In-Feed, and budgets on Series and Episodes were raised to maximum eligible spend to build subscription volume while the cost per subscriber held. Discovery looked inefficient on its own line, at $22 per subscriber in its first tracked week, and we kept it deliberately: Google confirmed it was feeding the Series and Episode audiences that converted far more cheaply, and its own cost per subscriber then fell to $8.40 the following week and $3.24 by late August. The program was not a one flight test. Episode support had reached Episodes 13, 14 and 15 by January, Episode 19 Shorts support ran in March, and the Vodcast was still pacing at 99 percent in the final pacing report of our tenure on September 2.

Results

Subscriptions are the result this program was managed to, and July is the peak month on record: 60,346 subscriptions from paid media, including 45,425 from the Series campaigns at $1.29 per subscriber and 13,767 from Episodes at $2.43. The efficiency improved from there: in the November reporting the Series line delivered 45.5K subscribers month to date at $0.92 per subscriber, with Episodes adding 7.1K at $2.49. The spring flight that built the audience ran at scale for very little: Series delivered 4.6 million views in April at a $0.02 cost per view, and Discovery had delivered 29 million impressions and 414K clicks at a $0.09 cost per click by the June 3 report. The monthly summaries agree: Series campaigns delivered 2.04 million views in March, 1.88 million in April and 2.29 million in May month to date, and Episode 2 drew 391,172 views in May. When Capital Group leadership asked for Episode 19 Shorts support in March with a goal of 4,000 views per version, each version hit its goal and was paused within days. Across July to November, actual Vodcast spend totaled $480,607 including platform fees, across the Brand Series, Episodes, Discovery and YouTube Audio lines.

We treated episodic video as a media product: one series destination, episode level promotion, audience splits on every campaign, and tracking on every route.

02 / ProShares

CPA discipline at scale, and AI Max on a short leash

The challenge

ProShares runs a paid search program of more than $9 million a year across geared, strategic, crypto and income funds, each with its own tickers and economics. In a program that size, the failure mode is quiet: spend drifts toward visibility, automation expands into whatever it can find, and acquisition cost stops being accountable. Our job as media agency of record is to keep every campaign family tied to a hard cost per acquisition goal while still testing what is new.

What we did

We run the account in campaign families by business line and objective, with explicit goals attached to each: a $30 CPA goal on CPA campaigns, $60 on conversion optimized campaigns, and separate visibility goals where the job is presence rather than acquisition. Pacing is reviewed weekly against goal, prior week and prior month, and budget moves toward the conversion drivers with the lowest CPAs. In October the pacing report closed the month at $221,822 of search spend against a $226,807 budget across the five business lines. When a fund had visibility campaigns only, as SMDV did, we proposed and built CPA campaigns alongside them, which the client approved. For AI Max, we worked with Google on candidate selection and an expansion view that ranked the campaigns with the most incremental conversion opportunity, we evaluated the AI Brief Asset Preview tool so AI generated copy can be checked against ProShares' own guidelines and guardrails before anything is enabled live, and we judge each AI Max expansion against its own pre-expansion CPA benchmark before it earns more scope. Candidacy itself is gated by our own criteria sheet: a campaign needs more than 10 percent of impression share lost to budget, at least 30 conversions a month, a tCPA or tROAS bidding strategy, and $50 to $100 or more of daily budget before AI Max is even considered. Google's Q4 expansion view sized the top 10 candidates at an estimated 937 incremental conversions per 28 days, which we treat as a sizing tool, not a promise.

Results

The AI Max expansions are passing that test. In the October 31 performance report, AI Max Oil cut CPA from a $19.38 benchmark to $15.65, a 24 percent improvement, and AI Max Crypto cut CPA from $17.29 to $14.88, a 16 percent improvement, while AI Max Broad Market held roughly flat. The same month, ProShares' share of clicks rose 52.7 percent month over month, with AI Max campaigns identified in our reporting as a key driver, finding long tail keywords the core campaigns did not reach. The CPA discipline underneath held at the same time: in our November performance report, Crypto CPA delivered a $12.78 CPA, improving 49 percent in a single week as Ether and Bitcoin conversions grew; Strategic CPA ran at $7.66; and Tactical CPA ran at $9.74, down from $21.01 the prior month, a 54 percent improvement led by the Nasdaq campaigns, all well inside the $30 goal.

Scale does not have to mean loose. We hold a search program of more than $9 million a year to CPA goals family by family, and we let AI expand only where the numbers say it has earned it.

03 / Sikorsky Credit Union

taking the lead on a local growth program

The challenge

Sikorsky Credit Union serves members across Fairfield, New Haven and Hartford counties in Connecticut, and grows one product at a time: free checking accounts, auto loans, mortgages and home equity, all inside a defined local footprint. When the credit union changed media leadership and handed the program to us with a clear brief, recorded in the onboarding project list: "Definitely want you to lead us with the plan, what we should do more/less of, any different tactics, etc." The program we inherited had momentum and a Covid reshaped budget. Our job was to keep what worked, sharpen the measurement, and aim the plan at the next products and markets.

What we did

We built the annual plan as one full funnel system, presented December 9, and flighted January 4 to December 31, at a $498,728 level (a $396,940 option was modeled alongside it). Five moves structured it. First, invest heaviest in the proven performers: paid search at $163,948 across Google and Bing, social, and retargeting. Second, expand video: YouTube rose from $9,200 of prior-year spend to $30,000, bought for tracked conversions, with CTV at $63,000 in spring and fall flights. Third, open a new awareness channel for younger members: $45,000 of streaming audio and podcasts aimed at the 24 to 34 audience. Fourth, flight campaigns to product seasonality, mortgage and auto pushes in their natural windows. Fifth, upgrade the discipline underneath: tracking, measurement and reporting cadence, including bi weekly insight updates on top of monthly reporting, and a success metric fitted to each tactic, CPA goals for search, social and retargeting, lift for CTV and audio, completed views and click through for YouTube, brand search trends and awareness surveys for radio.

Geography stayed the local edge. Targeting is built zip code by zip code around the branch communities, Milford, Stratford, Shelton, Seymour, Trumbull and Brookfield, with a geo fenced display test around Brookfield grocery locations, data targeting on in market bank and loan intenders, and WEBE radio ($46,280, 486 spots plus 246 Opening Bell reports) holding the local broadcast presence.

Results

The honest starting point is the baseline we were handed, and it was strong: in the program cut cost per new user 20 percent year over year, to $1.36, lifted visits to the Become A Member page by over 50 percent during the campaign, and reached 40.9 percent aided brand awareness, up from 31 percent in, with the second highest share of search among competing Connecticut credit unions. The plan was built to defend that efficiency while shifting mix toward the channels with headroom, and to report every channel against the metric that actually fits it.

Under our tenure the program reports in the same conversion language the framework set. In the channels delivered 1,745 conversions in total, led by Google Performance Max with 1,203, retargeting with 286, and location based mobile with 165, and the monthly key events rhythm follows the product calendar the plan flights to: peaks of 243 in July, 231 in April and 220 in June, the spring and summer lending and account season.

Local financial brands win on precision, not volume: the right zip codes, the right product intent, and reporting honest enough to move budget every week.

04 / KAYAK

We built KAYAK's programmatic program, and it returned 1.56x ROAS

The challenge

Travel intent does not arrive evenly. The person researching flights early in the morning and the person ready to book late in the evening behave like two different customers, and retargeting built at the day level treats them as one. KAYAK needed a programmatic program established from the ground up, one structured around when travelers actually research and book, not around calendar days.

What we did

We established KAYAK's programmatic campaigns and built the retargeting strategy at the daypart level, granular down to hours, not days. That structure came out of a simple discipline: match the message and the bid to the hour a traveler is most likely to act. We also ran dynamic ads for KAYAK, but only for specific location based searches, and only with extremely specific targeting, exclusions, and rules that the algorithm can follow. Automation did the scaling. Human rules set the boundaries it scaled inside.

Results

The programmatic campaigns we established for KAYAK yielded 1.56x ROAS. Just as important, the daypart structure became a repeatable lesson we still apply across travel accounts: the strongest campaigns are built with granular retargeting at the daypart level, down to hours, not days.

Travel is one of our deepest categories. Our experience includes Booking.com, KAYAK, Gulf Shore Tourism, Marriott Hotels, and Gaylord Hotels, and KAYAK is where our daypart discipline was proven.

05 / Catchmaster

off Amazon media that moved Amazon sales

The challenge

Catchmaster is a challenger pest control brand taking share from legacy chemical brands where the category is actually bought: Amazon, a $1.2B pest control category on the platform. Shelf position there is earned, not bought once. Best Seller Rank, organic keyword rank, and Buy Box position all move with traffic quality, so media run outside Amazon has to be judged by what it does inside Amazon.

What we did

We built a closed loop, off to on Amazon media engine tied to real life pest moments. Audience profiles were built from historical purchase velocity data across 14 target markets, then geo flighted to high converting DMAs in line with localized pest surges, including Chicago, Philadelphia, and Dallas Fort Worth. Product led short form video on Meta, competitor search conquesting, and off Amazon retargeting all routed to Amazon ASIN detail pages and the brand storefront, and the loop was closed with Teikametrics and Amazon Attribution across the off Amazon channels, so external spend could be read against Amazon outcomes market by market.

Results

Seasonal short form Meta video drove traffic straight to Amazon product pages at a $0.61 effective cost per click, with a click through rate 50 percent higher than static display. Search conquesting against competitor brands reached a click through rate of up to 9.60 percent, against a 1.65 percent eCommerce benchmark. And the compounding effect the engine was built for showed up in the outcome that matters: Catchmaster's Amazon sales growth outpaced the category benchmark by 4.8x, with $7.7M generated in the top focus markets, making it one of the fastest growing pest control brands on Amazon.

External traffic, measured properly, compounds: it buys the click today and improves the organic shelf the click lands on tomorrow.

Get in touch

Let's talk about your media.

Tell us what you want to improve, which channels you run and where you need support. We will talk through the fit, the scope and the next step. Brands and agency partners are both welcome.