AMA: Heap Vice President, Marketing Acquisition & Growth, Sheena Sharma on Campaign Operations

* Clear QA processes and approval processes. At Envoy we actually created a Google Form that at least one person (who didn't build the campaign/email) had to fill out to confirm that every critical QA element was covered (ex. subject line, correct links, correct targeting on emails). AND, we had a different person who was the official approver who could say the email was good to go AFTER they had seen the email QA come through. * Realistic SLAs for (1) acknowledging requests and (2) resolving or completing tickets. For example, tier your requests 1-4, with 1 being urgent and 4 being the least important/urgent. You need to build trust with your stakeholders by acknowledging and responding to high priority items, as well as showcasing that you've captured and will address lower priority items. BUT, you can't be fighting fires all day long. You have to prioritize long term initiatives as well. * A shared, regularly prioritized backlog. I have found that creating a shared and visible backlog of key projects helps your stakeholders understand what other asks the team has. You should aim to update and re prioritize the backlog every other week (no less frequently than monthly). AND, you should share it via email and in other key channels. * Both async and synchronous communication channels. Campaign operations can be tedius and repetitive, but it is important to not just rely on slacks or emails to communicate what the team is working on. Using regular meetings to share priorities and challenges is important to ensure that campaign operations processes are being followed, and that folks can give feedback. * Regular feedback loops. For example, if you roll out a new email QA process - you should set a meeting or feedback request out after 30 days, 60 days and 90 days of setting up the new process. It's really important to make sure that people trust that you will take their feedback into account and iterate on the process. If you don't do that, people will often jus

Typically, I've seen this fall under marketing operations. There have been cases where marketing operations owns more of the tech stack, integrations, and programs like scoring and database management, and campaign operations + execution sits within the demand generation teams. I've seen a decentralized model work well for global organizations, where you need coverage across different time zones and for different channel and campaign needs. In general, if you can have more than one person on Marketing Operations, I would recommend having one person focused on tech stack, and another focused on campaign operations. It can be hard to have one person responsible for both.

I think that strategically, there are definitely some things that make sense to work from a centralized perspective: Ex. global lead scoring, tech stack integrations, lead routing, website operations. Then, there are places where decentralized operations make sense, especially as it relates to regional demand generation programs that need to be localized - emails, events, field, campaign activities. It became not scalable for us at Eventbrite at a certain point to have a single email marketing manager who was supporting the globe, though that worked for us for a time. So I think there's also an argument for starting centralized with some functions until it just doesn't scale for your business.

I think A/B testing and evaluating statistical significance is obviously the highest fidelity type of testing you can do, but in the startup world I find that the speed with which you can test and learn is limited if you solely rely on A/B tests. I have found that thinking about things from an MVP perspective (minimum viable product/project) can help you learn and iterate more quickly. For example, if you are launching a new type of email program to see if triggered emails do better than traditional nurtures, taking an A/B testing approach might take you one or two quarters to see a significant lift. However, if you are just looking for any kind of initial signal, you could get learning much more quickly if you launch and then iterate on your initial results.

I have a three-pronged strategy that works pretty well: And that's the concept of (1) quarterly plans, (2) documented strategic briefs for every campaign and (3) quarterly retrospectives. I love the concept of quarterly retrospectives, where you revisit your strategy (which should always be clearly documented), and reflect on the performance of your campaign. You should include both quantitative results, as well as collecting qualitative feedback from internal stakeholders, as well as bringing in feedback from customers or prospects. My retrospectives usually follow a similar format: What did we say we would do, what did we actually do, and now what do we want to do differently going forward? You should time the retrospectives to be IN ADVANCE of the next quarter's planning, so that you can bring in the learnings from the last quarter into the next quarter's plan. What this looks like practically is that you are often already planning the Q+1 when you are reviewing the most recent results. For example: By January 15 you want to do the retro for Q4's email performance, covering October-December. If your Q2 starts on March 1, you probably want to have your quarterly plan set no later than Feb 1 (30 days before the quarter starts). That means you are using Q4 email results to inform Q2 strategies - but if you are doing this on a rolling basis each quarter you are always finding ways to iterate and improve on what you've done in the past. Finally, I think it's important to leverage lateral thinking and work to apply learnings from across the organization to your campaigns. For example, you should be paying attention to your competitive win rates and what's happening with win/loss reporting, as you can then bring those trends into your marketing campaigns for the next quarter.